Topics

Tek 422 ac/dc power supply issues

Renée
 

Hi everyone-
I am hoping someone may be able to assist in fixing the DC-DC converter of this cute scope.

I recently acquired a Tek 422 AC/DC that has ps issues. The history is it was supposedly working 20 yrs ago then put in storage. now it is mine..beautiful condition, cute scope! low 8k sn I would really like to get it back in use....especially on Batteries.......and no the batteries were not included.......thank heavens.
so far-
i checked all resistors/diodes/xstrs/caps in ps circuit as best as possible.
I have found :
In the DC -DC converter Q1184 (2n1899/151-163) is bad. fuse had not blown. (possibly not run long enough to cause it to go). I replaced both Q1184 and Q1174 with D1148 (what I thought to be a reasonable sub I had a matched pair from an old audio amp, I temporarily mounted them). NO Good draws too much current. Isn't this just a class B amp?

I cannot check the indicator without the supply...gee.....wish I had the AC supply too! (circuit looks soooo much easier!)

Maybe my sub was poor choice( too much Hfe)? Any one have a better suggestion ? I am thinking a FPJ13009 may be a better sub....much lower hfe.
I know wrong package I think I can make it look pretty and mount properly. how critical are the specs of these transistors?
the only list I could find is Motorola 1966 data book. show Hfe 10-30, Ib =1A, Ic=10A ( I can attach that page if needed)

I am thinking other possible subs other than FPJ13009 are MJE3055, BUV27, BUL98 these are available at Mouser as the original xstr seems unobtanium.

I managed to test the Start circuit, the feedback and -12 supply ( it makes -8V at C of Q1154). (using exterior ±12 to run the circuits).

Attached pwr per Calibration with 11.5V and it appears I have the oscillator functioning. The oscillator (~7 khz ) did appear to want to start and circuit wants to run? ( I had the D1148's installed and was afraid to let it run long for rear of trashing the transformer)
The big problem is too much current in the input circuit, 4.5A at 12V input appears my sub for Q1174/84 is incorrect...they get really hot. and of course as input V rises so does current.( yes I finally popped the fuse) it should not go over 2.5A? (3A fuse). without the Q1174/84 circuit draws what seems to be reasonable (very little)

questions-
is there a problem with testing just the pwr sply section by defeating the DC interlock (pins 13-24)?

are the C1170/71 (80 uf caps) a feedback network in the circuit? ( i measured 1.5ma leakage each), these were replaced with a singe 150uf/100V good cap.

the start circuit appears to be good. but one needs the osc to generate the ±12 supplies that run the osc that drives the ........well big loop.

Other than the Q1174/84 issue, something is obviously wrong ( I removed the D1148's), suggestions are appreciated.

I really need some help, loops are a pain.
Thank you in advance.
Renée

 

First of all this reply will regurgitate the thread to the top of the heap. Easier to find that way.

As I said in the 7904 thread I just started looking at this power supply.

My book shows the 2N1899 as Vcb 140 (!), Vcb 50, Veb 5, Ic 10, Pd 125, Hfe 20 @ 10 mA, GBP 50 MHz. The 2SD1148 is good in most parameters except ft, which is only 20 MHz. it may not be switching fast enough which means heat. My book shows the MJE3055 as having GBP of only 2 MHz so that's out. The BUV27-D has a turnon time of 400 nS, which may mean it needs more base current despite the higher hfe, and it drops 750 mV when they consider it on at 1 amp. That is wasted power, You know where that goes. Looks like the FPJ13009 is fast enough but the gain might not cut it. The BUL98 looks fast enough but again has lower gain. Some of these possibilities you mentioned are much higher voltage. There is always a tradeoff. Running them at a much lower voltage might bring in a higher capacitance between C and B, which is not good. So it appears your basic parameters are met, but those switching characteristics might be in the way. I'll see what I can find. You are most likely right in that these replacement transistors just don't cut it. and the fact that it runs means the likelihood of a an overload on the secondary is not that high.

You are in very little danger of damaging the transformer, but switching transistors get thermal runaway just like audio outputs. It's just in this case they get hotter which slows them down and makes the switching time longer which is when they dissipate power which makes them hotter which slows the switching time which...

So you are probably on the right track about the replacements not being up to snuff. I am sure I can find something.

Now on mine, I assume thermal because the case of the power supply section gets warm. Those transistors are not exactly coupled thermally to the case so they must be much hotter than the case temperature lets on.

But I see no viable thermal cutoff or even current limiting, so the shutdown is a mystery since there is no permanent damage. Depending on the mounting arrangement of the choppers, it could conceivably be a bad solder connection but I am not taking that to the bank, i need a bit of time to study it a bit. I have a few other things going on and I might have to put it aside for a short time. I need to look at someone's printer/fax, I have upcoming appointments with a cardiologist and an ophthalmologist, and a court case in which I will be operating pro se. The pretrial for that comes up Wednesday and I am getting my medical records for my defense. They might just dismiss it because I have mounds of evidence and the officer never even signed the complaint, plus I can prove his assertions were in error, or at least reasonable doubt of same. They will learn to not screw with me and I am not a revenue source like most sheeple. I actually had the last city in which I lived trained, the cops would not even set foot on my property. They caught me with a suspended license, beer between my legs and all that and just gave me a ticket, and if I had a legal river with me they wouldn't even tow my car. So it is time for a lesson for these folks here. I am a bit preoccupied with that these days, and getting my eyesight fixed and it is not a matter of gasses. My eyes need about ¼ million bucks worth of work so I can work gainfully again. If I can do that fine, if not I guess I will be filing for disability. So I do take some time off the bench, bear with me.

I don't feel like it right now, but Digikey has an extremely good selection system for parts. I can put in almost all of the parameters and it will give me all the options that they can get their hands on. Hopefully it is not a drop ship item with a minimum order up in the thousands. Some are, but they deal with more than individuals, like engineering departments and such. Then again I think you might be able to restrict the search to minimum orders of one or two. Their table of parameters spans more than almost any screen, there is a scroll at the bottom to find the other ones. And they have current datasheets on damnear everything as long as it is there. So we will find those transistors. in my case, I will have to make sure mine are original, and then find out how the hell it could be shutting down without damage. You almost certainly simply need better transistors. They are out there.

More later. Do RSVP with any new information I have not gotten yet.

This is not bad, in studying yours I could gain some insight into mine. So no sweat.

 

Oops, Vce 50

 

On Sun, Apr 8, 2018 at 06:23 pm, Renée wrote:


questions-
is there a problem with testing just the pwr sply section by defeating the DC
interlock (pins 13-24)?
You should load the output first. Like I told the guy with the 7904, a light bulb should do it. It would probably be best on the 55 volt line.

"are the C1170/71 (80 uf caps) a feedback network in the circuit? ( i measured 1.5ma leakage each), these were replaced with a singe 150uf/100V good cap."

My copy shows 1171 & 1172 as 80 uF, 1170 is a 0.1 uF. They are not part of the feedback loop, all they do is establish a common AC between the emitters of the choppers and the 5 uH and diode going to the primary of T 1201. That impresses a positive spike on the emitters of the choppers when the spikes come from the collectors, probably improves effective turnoff time.

The start circuit is not a loop per se. It takes voltage right from the input and regulates it somewhat. When the voltage comes along out of D 1242 & 1243 the emitter of Q 1183 is moire positive than the base which puts it in cutoff. The feedback is compared and amplified by Q 1134, 144 & 1154 and sent to the actual control circuit with the tunnel diode. I am not good with tunnel diodes but there is likely to be no problem there. Just don't blow it, you'll probably never find a replacement for that diode. Tek knows how to use tunnel diodes, I don't. The one in the triggering circuit seems to make it truly superior, I had my old one locking on noise. Audio and composite video no problem, and those give most scopes a hard time. I don't see how it is any better in the PS there but since that's that way it is I would try to leave it alone.

If your choppers are getting hot they are obviously being driven. The primary of T 1171 only gets about 3.5 volts, it might be a good idea to make sure it is getting it. If there is no overload on the secondaries of T 1201, the only reason for the choppers to get hot is that thy are not switching fast enough. There is no switching enhancement, like resistors and caps to the bases so it is must be the transistors themselves.

Those TO-81 cases are not common, you might have trouble finding something that'll fit. Probably the best thing to do is figure out what available case style is easiest to adapt and confine a Digikey search to that case, and then go for the other parameters. But what case ?

Do you have access to a small milling machine, or at least a drill press with a cross slide vise ? With that you might be able to modify the mounting (or make an adapter) for a modern case style. I know the rule is to modify the part not the unit but that is only when possible. Or, it may be possible to make a TO-220 fit in there if one can be found that meets all the other specs. I don't remember any TO-220 that can dissipate 125 watts. Maybe a TO-202 but that is bigger and could make more problems.

I am going to have to get into the rear part of the PS in mine, I'll have a look. If it's impossible all that means is it takes longer.

Renée
 

yes I goofed on the C1170/1/2 numbers. thank for the analysis.

after reading the post re the 7904 I realized that yes I do need to load the supply.....so it is now always loaded either with the head or some other resistive load (need to build).

no problem with the xstr package, I can mill what ever I need.

MJE1661 is what TEK was appariently putting in

On 2018-04-12 03:33 AM, Jeff Urban wrote:
On Sun, Apr 8, 2018 at 06:23 pm, Renée wrote:

questions-
is there a problem with testing just the pwr sply section by defeating the DC
interlock (pins 13-24)?
You should load the output first. Like I told the guy with the 7904, a light bulb should do it. It would probably be best on the 55 volt line.

"are the C1170/71 (80 uf caps) a feedback network in the circuit? ( i measured 1.5ma leakage each), these were replaced with a singe 150uf/100V good cap."

My copy shows 1171 & 1172 as 80 uF, 1170 is a 0.1 uF. They are not part of the feedback loop, all they do is establish a common AC between the emitters of the choppers and the 5 uH and diode going to the primary of T 1201. That impresses a positive spike on the emitters of the choppers when the spikes come from the collectors, probably improves effective turnoff time.

The start circuit is not a loop per se. It takes voltage right from the input and regulates it somewhat. When the voltage comes along out of D 1242 & 1243 the emitter of Q 1183 is moire positive than the base which puts it in cutoff. The feedback is compared and amplified by Q 1134, 144 & 1154 and sent to the actual control circuit with the tunnel diode. I am not good with tunnel diodes but there is likely to be no problem there. Just don't blow it, you'll probably never find a replacement for that diode. Tek knows how to use tunnel diodes, I don't. The one in the triggering circuit seems to make it truly superior, I had my old one locking on noise. Audio and composite video no problem, and those give most scopes a hard time. I don't see how it is any better in the PS there but since that's that way it is I would try to leave it alone.

If your choppers are getting hot they are obviously being driven. The primary of T 1171 only gets about 3.5 volts, it might be a good idea to make sure it is getting it. If there is no overload on the secondaries of T 1201, the only reason for the choppers to get hot is that thy are not switching fast enough. There is no switching enhancement, like resistors and caps to the bases so it is must be the transistors themselves.

Those TO-81 cases are not common, you might have trouble finding something that'll fit. Probably the best thing to do is figure out what available case style is easiest to adapt and confine a Digikey search to that case, and then go for the other parameters. But what case ?

Do you have access to a small milling machine, or at least a drill press with a cross slide vise ? With that you might be able to modify the mounting (or make an adapter) for a modern case style. I know the rule is to modify the part not the unit but that is only when possible. Or, it may be possible to make a TO-220 fit in there if one can be found that meets all the other specs. I don't remember any TO-220 that can dissipate 125 watts. Maybe a TO-202 but that is bigger and could make more problems.

I am going to have to get into the rear part of the PS in mine, I'll have a look. If it's impossible all that means is it takes longer.


Renée
 

Hi Jeff-
yes I goofed on the C1170/1/2 numbers. thank for the analysis.

after reading the post re the 7904 I realized that yes I do need to load the supply.....so it is now always loaded either with the head or some other resistive load (need to build).

no problem with the xstr package, I can mill what ever I need. I think the TO-3 package the most difficult to adapt in terms of having to disassemble the two halves and place new holes, however most likely the best heat dissipation. although I find the heat sink pathetic at best meaning it should not be generating much heat.....

MJE1661 is what TEK was apparently put in per John Sykes who sent me a picture of the thing mounted almost like I had done mine for a no holes approach. From John-  " Someone (I assume Tek) had replaced *just one* of the original devices with a TO127 device, 151 - 0469  = MJE1661" ...
I see the thermal xfer as atrocious being mounted on component side of the board, again not much heat generated.
the good thing is that when this is done a solution will be available for the future......

Renée

On 2018-04-12 03:33 AM, Jeff Urban wrote:
On Sun, Apr 8, 2018 at 06:23 pm, Renée wrote:

questions-
is there a problem with testing just the pwr sply section by defeating the DC
interlock (pins 13-24)?
You should load the output first. Like I told the guy with the 7904, a light bulb should do it. It would probably be best on the 55 volt line.

"are the C1170/71 (80 uf caps) a feedback network in the circuit? ( i measured 1.5ma leakage each), these were replaced with a singe 150uf/100V good cap."

My copy shows 1171 & 1172 as 80 uF, 1170 is a 0.1 uF. They are not part of the feedback loop, all they do is establish a common AC between the emitters of the choppers and the 5 uH and diode going to the primary of T 1201. That impresses a positive spike on the emitters of the choppers when the spikes come from the collectors, probably improves effective turnoff time.

The start circuit is not a loop per se. It takes voltage right from the input and regulates it somewhat. When the voltage comes along out of D 1242 & 1243 the emitter of Q 1183 is moire positive than the base which puts it in cutoff. The feedback is compared and amplified by Q 1134, 144 & 1154 and sent to the actual control circuit with the tunnel diode. I am not good with tunnel diodes but there is likely to be no problem there. Just don't blow it, you'll probably never find a replacement for that diode. Tek knows how to use tunnel diodes, I don't. The one in the triggering circuit seems to make it truly superior, I had my old one locking on noise. Audio and composite video no problem, and those give most scopes a hard time. I don't see how it is any better in the PS there but since that's that way it is I would try to leave it alone.

If your choppers are getting hot they are obviously being driven. The primary of T 1171 only gets about 3.5 volts, it might be a good idea to make sure it is getting it. If there is no overload on the secondaries of T 1201, the only reason for the choppers to get hot is that thy are not switching fast enough. There is no switching enhancement, like resistors and caps to the bases so it is must be the transistors themselves.

Those TO-81 cases are not common, you might have trouble finding something that'll fit. Probably the best thing to do is figure out what available case style is easiest to adapt and confine a Digikey search to that case, and then go for the other parameters. But what case ?

Do you have access to a small milling machine, or at least a drill press with a cross slide vise ? With that you might be able to modify the mounting (or make an adapter) for a modern case style. I know the rule is to modify the part not the unit but that is only when possible. Or, it may be possible to make a TO-220 fit in there if one can be found that meets all the other specs. I don't remember any TO-220 that can dissipate 125 watts. Maybe a TO-202 but that is bigger and could make more problems.

I am going to have to get into the rear part of the PS in mine, I'll have a look. If it's impossible all that means is it takes longer.


 

With only a 3 MHz ft ? Well, if you trust the source it is worth a try if you can get one cheap enough, well two actually.

Can you forward that picture ? I see the transistor is a TO-225AB case but don't see how it would be mounted. but then even if it doesn't work right it means that case style will fit which is good to know.

Renée
 

Hi Jeff-
Yes I was questioning ft as well however...he says it works .... lets see if I can get the pic copied here...yes!!! see below
I found that I can fit a TO-3 from the underside too! ....I figure I can machine something for the heat xfer for this style of case.

On 2018-04-12 07:47 PM, Jeff Urban wrote:
With only a 3 MHz ft ? Well, if you trust the source it is worth a try if you can get one cheap enough, well two actually.

Can you forward that picture ? I see the transistor is a TO-225AB case but don't see how it would be mounted. but then even if it doesn't work right it means that case style will fit which is good to know.


Renée
 

guess no picture I will send direct.
Renée

Mark Wendt
 

On 04/13/2018 02:36 AM, Renée wrote:
guess no picture I will send direct.
Renée
This list is set up to not allow attachments.  Post the image in the Photo section and then post the link to it.  Hopefully, groups.io will get the search functionality up and running for the photo section soon.

Mark

Renée
 

I Think I made an album https://groups.io/g/TekScopes/album?id=44362..how do i make it a link? ..hmmm..learning how to add/make picts....should have renamed them.

Shows has measurement of the base and collector outputs with a different Sub FPJ13007, so the oscillator is working and I can change the Freq slightly, Use the last 3 digits- Explanation both 50us -Collector #824 but is 2V scale, Base #318 0.5 V scale...so not much gain. #321 mounting of the xstrs.
questions? yes, not sure where to go from here.

Renée
 

Renée
 

I realized the other item somewhat important input Voltage is 11.5....drawing 3.0A.
it is item #2 in the calibration section 6-4 Adjust Blocking Osc freq. way low in output even at this voltage!
sorry for my inability to get the posts correct......still figuring things out slowly.....
Renée

 

Yes, it says 24.5 coming out of the main power transformer. You got a bit less than half that. Such a low voltage MIGHT cause it to run hot so I would check the 1,300 uF caps, ripple and whatever and see if they're getting hot.

Also, if you are draining 3 amps at 11.5 that is around 30 watts. the plate on back says 27 and that figure includes all the inefficiency introduced by the iron and rectifiers. That may explain some of the heat generated as well.

If the scope works otherwise, it is unlikely anything is shorted in the "display unit". IF Q 1033 is shorted and the battery pack is at zero volts it would pull about ¼ amp, that's about 10 watts. Depends on where you measured that 3 amps. Was the charging regulator included or not ? Even if not, the low voltage could still warm up those choppers. If you can get to R 1033 those voltages might prove interesting.

I'll see if I can check all that on mine because it looks like we have the same problem. Possibly your thermal cutout operates on the high side and mine on the low side of 155F, or yours could have been run more with people waiting around for it to come back on and then use it until it cycles off again. You never know, it may have been the only scope on site and they needed it. Sometimes necessity is the Mother of abuse.

If I don't have that high current in mine I might figure out a way to see what the display unit is pulling off the sources. Not sure how, I don't have a Hall effect magnetic current probe, my AC one is designed for 60 Hz, and I am not desoldering all those supplies to insert a 0.1 ohm or whatever to measure it that way. I'll see what I can come up with. however, if I DO have excessive current then it is not worth it, any readings would be useless.

I'll probably be back on it tonight but no guarantees.

Renée
 

actually I measured the current again at 11.5V it is closer to 5A. (the range switch on amp meter is intermittent) this is the procedure for adjusting the Blocking Osc Freq section 6-4 in the manual.
one applies 11.5V at the battery terminals, and takes a measurement, do the adjustment so it looks like the book.
the other thing is the attack time for this particular xstr (FPJ13007) seems to be really fast , so no leading edge prior switch.
I will add a couple more photos....figured out how to rename pictures with scope settings.
obviously the xstr is wrong.....I am wondering if I can add an emitter R or lower the input by swamping the Base somehow to slow the turn on time?

Being you are having a similar issue that will lead to Q1184 failure due to heat...the internal junctions on mine are so leaky not even act like diodes...even the EC junction is sort of shorted.....

Renée
 

I have been testing/running the Dc-DC converter with an external pwr supply and connecting to the battery terminals as to be able to monitor the current and control the voltage, you might try the same. So no charging is occurring. (I did test the charge circuit per book- it is ok)
I had already checked the ripple for the onboard ac-Dc supply it is good.
on to yours-
so why is yours getting hot? check the waveform at D1176/7 see if it matches the book with the procedure 6.4 ( 11.5V input) if the Freq is off then that may be the cause.....possibly the junctions of the choppers are going south due to the heat and will only get worse due to heat....big circle here....

I also found in mine why the fuse did not pop with my initial trials on AC someone had reversed the Ac and DC fuses....I goofed and only removed one at a time and saw they were good and put'em back.....maybe that necessity abuse cycling issue. who knows as the previous owner is no longer with us...
I have never had the actual scope (indicator portion) working So...I have no idea if the indicator section functions or not. I took a bunch of R measurements to make sure no shorts both sides of the connector......HV section excluded as it is just xfmr coupled.

On 2018-04-18 09:05 AM, Jeff Urban wrote:
Yes, it says 24.5 coming out of the main power transformer. You got a bit less than half that. Such a low voltage MIGHT cause it to run hot so I would check the 1,300 uF caps, ripple and whatever and see if they're getting hot.

Also, if you are draining 3 amps at 11.5 that is around 30 watts. the plate on back says 27 and that figure includes all the inefficiency introduced by the iron and rectifiers. That may explain some of the heat generated as well.

If the scope works otherwise, it is unlikely anything is shorted in the "display unit". IF Q 1033 is shorted and the battery pack is at zero volts it would pull about ¼ amp, that's about 10 watts. Depends on where you measured that 3 amps. Was the charging regulator included or not ? Even if not, the low voltage could still warm up those choppers. If you can get to R 1033 those voltages might prove interesting.

I'll see if I can check all that on mine because it looks like we have the same problem. Possibly your thermal cutout operates on the high side and mine on the low side of 155F, or yours could have been run more with people waiting around for it to come back on and then use it until it cycles off again. You never know, it may have been the only scope on site and they needed it. Sometimes necessity is the Mother of abuse.

If I don't have that high current in mine I might figure out a way to see what the display unit is pulling off the sources. Not sure how, I don't have a Hall effect magnetic current probe, my AC one is designed for 60 Hz, and I am not desoldering all those supplies to insert a 0.1 ohm or whatever to measure it that way. I'll see what I can come up with. however, if I DO have excessive current then it is not worth it, any readings would be useless.

I'll probably be back on it tonight but no guarantees.


 

Remember I said no promises about tonight ? The reason ? HA HA, I got an old PS/2 keyboard on the PC down there now and finally got it back online. So I want to use that one when working on stuff. The letters are worn off of it and I have to paint them. I can't use a Sharpie because it is black. LOL. But really, it goes back up tonight and I will check a few things.

I appreciate your information. So the DC convertor is actually pulling that 5 amps. OK. My small bench PS only does 2 amps, my big bench PS will need the variac to regulate and I'll have to dig up a resistor to measure current. It goes up to 100 volts at 5+ amps. And you are running it without the display unit and it is pulling 5 amps, Hmm, That's almost 60 watts. It is literally wasting more power than it is using. I'll try mine with the display first and see what I get.

You don't want to slow the switching time. It is not like a delay then it switches BAM like that. It is the switching itself takes more time and during that time there is voltage and current simultaneously so therefore power is dissipated. The longer that time the more average power which turns into heat. The other sources of heat are leakage current when it is turned off and Vce voltage drop when it is turned on. The leakage is usually negligent, especially at these lower voltages. The Vce sat is another story. If the transistor is turned on fully that voltage could be 0.6 V, so for a round number take ½ volt. At 1 amp that is ½ watt. at 5 amps it is 2.5 watts. If it is not turned on hard enough, it could easily be 1 volt which doubles all those numbers. Even though the transistors may be capable of dissipating over 100 watts it is still limited by the heat sinking. the power rating must be derated,

So all that plus the fact that you have a circuit board between the transistor and the heatsink I would say it is suffering and about due to fail.

I'll be back.

 

Started out it was draining 2.9 amps, now after adjustment 2.5. One the -12 was readjusted all the other voltages came into line. I don't know how important it is but to get the proper 12 volts the control was at the end of its range. Not its physical range, a little over 12 and it would turn more but there was no further increase.

That "approx". 94 uS and 140 uS (which is not exactly what their waveform shows anyway), doesn't come in perfectly, one or the other is off. I compromised. "Approx" is certainly the right word.

This is all with the display unit connected and running. They did not specify otherwise so that's how I did it.

I did not run it all that long, but there didn't seem to be any significant heat buildup. Next time, maybe later tonight I will run it longer, maybe occupy myself with something else for an hour or so.

"...the internal junctions on mine are so leaky not even act like diodes...even the EC junction is sort of shorted....."

If you're reading them in circuit they will read 7 different kinds of wrong.

"...actually I measured the current again at 11.5V it is closer to 5A..."

Typically I do not trust ammeters. I use a low value resistor and do the math. In this case I had a nice 0.15 ohm. There was 438 mV across it so that means 2920 mA. At the end it was like 370 something.

Ideally you get a 0.1, then the math is super simple, just move the decimal place. No blowing the meter if something shorts, and it is as accurate as the resistor and your voltmeter.

If the heat doesn't build up I will put it back on its own 24 volts and see what happens. The frequency might just be that critical.

Renée
 

Hi Jeff-
no biggie on doing things tonight...I leave tomorrow afternoon for work (performances over the weekend )..
how about a silver sharpie for the letters......works great on black things..label wall warts etc....
I should be making reeds...but no, I had time to look at the signals with a borrowed scope....oh well....
I began with the display removed...then the 7904 discussion ...hence always with the display connected from now on.
heat issue- yes as I said in the beginning that this method of mounting is only temporary as I do not like the thermal xfer thru the board....questionable at best.
thanks for the info on switching.
Renée

On 2018-04-18 04:49 PM, Jeff Urban wrote:
Remember I said no promises about tonight ? The reason ? HA HA, I got an old PS/2 keyboard on the PC down there now and finally got it back online. So I want to use that one when working on stuff. The letters are worn off of it and I have to paint them. I can't use a Sharpie because it is black. LOL. But really, it goes back up tonight and I will check a few things.

I appreciate your information. So the DC convertor is actually pulling that 5 amps. OK. My small bench PS only does 2 amps, my big bench PS will need the variac to regulate and I'll have to dig up a resistor to measure current. It goes up to 100 volts at 5+ amps. And you are running it without the display unit and it is pulling 5 amps, Hmm, That's almost 60 watts. It is literally wasting more power than it is using. I'll try mine with the display first and see what I get.

You don't want to slow the switching time. It is not like a delay then it switches BAM like that. It is the switching itself takes more time and during that time there is voltage and current simultaneously so therefore power is dissipated. The longer that time the more average power which turns into heat. The other sources of heat are leakage current when it is turned off and Vce voltage drop when it is turned on. The leakage is usually negligent, especially at these lower voltages. The Vce sat is another story. If the transistor is turned on fully that voltage could be 0.6 V, so for a round number take ½ volt. At 1 amp that is ½ watt. at 5 amps it is 2.5 watts. If it is not turned on hard enough, it could easily be 1 volt which doubles all those numbers. Even though the transistors may be capable of dissipating over 100 watts it is still limited by the heat sinking. the power rating must be derated,

So all that plus the fact that you have a circuit board between the transistor and the heatsink I would say it is suffering and about due to fail.

I'll be back.


Renée
 

oh more tonight.....cool....
the internal junctions of Q1184 were measured with device totally out of circuit...I agree, no way to do while connected.
I will If I have time to use a R of .1 or thereabouts, I should have verified the metre, and you are entirely correct Amp meters can be an issue.
The scope display pictures are all with 11.5V input to the Batt terminals. I have not tried to run it at 24/28V...my supply will not handle it...I have to repair the other one....oops...what i get for taking in strays......

From what I read and understand the -12V sets the pulse width so between that and the Freq could be all that is needed to eliminate your heat issue. Did you per chance photograph the 11.5V input display of the Freq at D1176/7? or is the book picture close?
The one thing in the "book" picture is the leading edge...I do not have a leading edge. This may be related to the devices I shoved into service are still incorrectly chosen ? possibly too fast?
Hope you heat issue is solved.
Renée

On 2018-04-18 09:18 PM, Jeff Urban wrote:
Started out it was draining 2.9 amps, now after adjustment 2.5. One the -12 was readjusted all the other voltages came into line. I don't know how important it is but to get the proper 12 volts the control was at the end of its range. Not its physical range, a little over 12 and it would turn more but there was no further increase.

That "approx". 94 uS and 140 uS (which is not exactly what their waveform shows anyway), doesn't come in perfectly, one or the other is off. I compromised. "Approx" is certainly the right word.

This is all with the display unit connected and running. They did not specify otherwise so that's how I did it.

I did not run it all that long, but there didn't seem to be any significant heat buildup. Next time, maybe later tonight I will run it longer, maybe occupy myself with something else for an hour or so.

"...the internal junctions on mine are so leaky not even act like diodes...even the EC junction is sort of shorted....."

If you're reading them in circuit they will read 7 different kinds of wrong.

"...actually I measured the current again at 11.5V it is closer to 5A..."

Typically I do not trust ammeters. I use a low value resistor and do the math. In this case I had a nice 0.15 ohm. There was 438 mV across it so that means 2920 mA. At the end it was like 370 something.

Ideally you get a 0.1, then the math is super simple, just move the decimal place. No blowing the meter if something shorts, and it is as accurate as the resistor and your voltmeter.

If the heat doesn't build up I will put it back on its own 24 volts and see what happens. The frequency might just be that critical.