Topics

RD16HHF1 power curve flattening...some

K9HZ <bill@...>
 

Glenn this is a very good idea… 

 

There are a number of ways the existing driver stages could be cleaned up. 

 

A little compensation at higher frequency should get that 2dB back.  Also I’ve implemented an electronic attenuator at the first stage (fixed parameter driven off the band selection of course) to make the gain exactly flat over the entire frequency range.   Myself, I’m interested in going up to 54 MHz but that should not be too much of a problem for that type 43 material. 

 

 

Dr. William J. Schmidt - K9HZ J68HZ 8P6HK ZF2HZ PJ4/K9HZ VP5/K9HZ PJ2/K9HZ

 

Owner - Operator

Big Signal Ranch – K9ZC

Staunton, Illinois

 

Owner – Operator

Villa Grand Piton – J68HZ

Soufriere, St. Lucia W.I.

Rent it: www.VillaGrandPiton.com

Like us on Facebook! facebook icon

 

Moderator – North American QRO Yahoo Group.

 

email:  bill@...

 

 

From: BITX20@groups.io [mailto:BITX20@groups.io] On Behalf Of Glenn
Sent: Saturday, March 10, 2018 8:26 PM
To: BITX20@groups.io
Subject: Re: [BITX20] RD16HHF1 power curve flattening...some

 

I have been experimenting on a different tack. Imnproving the response of the PA driver stages.

I used MPSH10's in my test board build and FT37-43 toroids wound as either trifilar or bifilar according to stage needs, of 10 turns.

Comparing the original uBITX driver which has a ~7db less gain at 10M end, this version has ~2dB drop.

I don't have any spare output FETS or I would check that out also, driven by the this test board.

glenn
vk3pe


Virus-free. www.avg.com

Glenn
 

Hi Bill, yes  a bit more work should get the 2dB back also. Just needs 'time'. I didn't know what spec toroids were used in the original, but had the FT37-43 on hand. The MPSH10 was on hand also, a better FT than the 2N3904's.  I did very briefly play with the bias also but  no real gains there.

Once it's all flattened out it would simplify fitting of some sort of power control immensely.

glenn

M Garza <mgarza896@...>
 

Nice Work, Glenn.

It looks like the MPSH10 is obsolete.  Doing some searching and comparing, the KSP10 is an exact replacement.   
It is in stock at Tayda, Mouser, and Digikey.  I only checked these three vendors.

What size wire did you use for your toroids?

Thanks.

Marco- KG5PRT



On Sat, Mar 10, 2018 at 8:52 PM, Glenn <glennp@...> wrote:
Hi Bill, yes  a bit more work should get the 2dB back also. Just needs 'time'. I didn't know what spec toroids were used in the original, but had the FT37-43 on hand. The MPSH10 was on hand also, a better FT than the 2N3904's.  I did very briefly play with the bias also but  no real gains there.

Once it's all flattened out it would simplify fitting of some sort of power control immensely.

glenn


John Backo
 

Dan's has MPSH10's/ 12/$US3.00.

I have found he has generally good transistors,
mostly old stuff. These are Fairchild pieces.

I think BGMicro has some a while back...but they
are no longer advertising them, so they are out.


john
AD5YE

Glenn
 

Plenty of MPSH10's on eBay.

I used 0.25mm wire  on the toroids. Just what i had basically.

glenn
vk3pe

M Garza <mgarza896@...>
 

Glenn,
Thanks for the wire size.
I usually do not buy semiconductors from Ebay.  I have been burned too many times.  Just my preference.

Thanks again,

Marco  - KG5PRT 

On Mar 11, 2018 5:02 AM, "Glenn" <glennp@...> wrote:
Plenty of MPSH10's on eBay.

I used 0.25mm wire  on the toroids. Just what i had basically.

glenn
vk3pe

John <passionfruit88@...>
 

Hello Glen,

Very interesting. Do you have measurements of you final's power across the bands by any chance?

All the best,

73, John  (vk2eta)

Glenn
 

Hi John

Totally out of the box uBITX, (ie no mods)  I checked power out in CW mode

120118

Checked using CW key at 12.5v dc in. VR1 at ~60% anti-clockwise. (factory setting)

·     

·         80M = 12w

·         40M = 8.5W

·         20M = 4.8W

·         10M = 1.2W

Turning up VR1 to full rotation.

80M = 15w
40M = 12w
20M = 7.8w
10M = 2w

glenn

Glenn
 

Marco

I bought the MPSH10's about 5-6years ago. Perhaps less of an issue with fakes then.

glenn

馬玉麟 <patrickylma@...>
 

Thanks for your information!

73 de BV5DD


2018-03-12 6:15 GMT+08:00 Glenn <glennp@...>:

Hi John

Totally out of the box uBITX, (ie no mods)  I checked power out in CW mode

120118

Checked using CW key at 12.5v dc in. VR1 at ~60% anti-clockwise. (factory setting)

·     

·         80M = 12w

·         40M = 8.5W

·         20M = 4.8W

·         10M = 1.2W

Turning up VR1 to full rotation.

80M = 15w
40M = 12w
20M = 7.8w
10M = 2w

glenn


Nick VK4PP
 

Well, i have dont my mods and powered up. Results are mixed:
Here are the mods.
RD16 replace IRF510.
New output transformer 1:4 BN43-3321 with a 330pf cap on input.
feed back resistors. from 22ohm to 1 K ohm
piggy backed R87 and R88 with 220pf capacitor.
RV1 full drive.

So results.
80M - 20W + ( blew my 2 Amp fuse)
40M - 18W - Seems stable.
30M - 17W BUT I get BAD feedback/ oscillation by just going PPT with no voice. The watt meter goes up to 18W!
20M - 15W  -  Same feedback issue
15M -  Same feedback issue.
10M -  Same feedback issue.

I have been thinking ans looking at other RD16 schematics, in alot of cases they dont use the feedback on the finals?
Also could this issue be cuase in the driver stage?

THanks guys, I am in over my head here....

73
10M - Same feedback issue

Nick VK4PP
 

Also,  RV1 Drive level has no effect on the feedback issues, happens just the same at 5W as at 15W...

Glenn
 

I completed my mockup of the PA driver stages.

as previous post:
  • used MPSH10's in all stages. 
  • Used FT37-43 toroids all wound with 10 turns 0.25mm wire either tri- or bi-filar according to original Sch.
  • Bypassed all emitter resistors with a 220pF cap.
  • swept from C80 input to T10 secondary (as wound)

Result is a very flat response. Better than 1db from 3Mhz to 54MHz.

glenn
vk3pe

M Garza <mgarza896@...>
 

Glenn,
Very nice work.  Thank you for sharing this info.

Marco - KG5PRT 

On Mar 11, 2018 11:02 PM, "Glenn" <glennp@...> wrote:
I completed my mockup of the PA driver stages.

as previous post:
  • used MPSH10's in all stages. 
  • Used FT37-43 toroids all wound with 10 turns 0.25mm wire either tri- or bi-filar according to original Sch.
  • Bypassed all emitter resistors with a 220pF cap.
  • swept from C80 input to T10 secondary (as wound)

Result is a very flat response. Better than 1db from 3Mhz to 54MHz.

glenn
vk3pe

William R Maxwell
 

Great work, Glenn.

Bill, VK7MX/3


On 12/03/2018 3:02 PM, Glenn wrote:
I completed my mockup of the PA driver stages.

as previous post:
  • used MPSH10's in all stages. 
  • Used FT37-43 toroids all wound with 10 turns 0.25mm wire either tri- or bi-filar according to original Sch.
  • Bypassed all emitter resistors with a 220pF cap.
  • swept from C80 input to T10 secondary (as wound)

Result is a very flat response. Better than 1db from 3Mhz to 54MHz.

glenn
vk3pe

Mharpen@...
 

Let me ask a stupid question. The feedback in the PA is negative isn’t it? Thus reducing it would encourage oscillations. Or am all wrong?

Mike n6cmy

K9HZ <bill@...>
 

Nik... looks like you have most of it right.  Plenty of gain, now just need to make it stable.  You might go to 15 meters and set the drive pot to zero.  Its gotta be stable with no drive or it wont work.  The model simulations don't show oscillations but the modeling parameters are not perfect for those transistors.  Probably need a little work on the feedback circuit to stabilize the finals, then it will be ok.  Again this is really good work and you are very close to a really good PA solution. 


Dr. William J. Schmidt - K9HZ J68HZ 8P6HK ZF2HZ PJ4/K9HZ VP5/K9HZ PJ2/K9HZ

 

Owner - Operator

Big Signal Ranch – K9ZC

Staunton, Illinois

 

Owner – Operator

Villa Grand Piton - J68HZ

Soufriere, St. Lucia W.I.

Rent it: www.VillaGrandPiton.com


email:  bill@...

 


On Mar 11, 2018, at 6:54 PM, Nik VK4PLN <nickpullen@...> wrote:

Well, i have dont my mods and powered up. Results are mixed:
Here are the mods.
RD16 replace IRF510.
New output transformer 1:4 BN43-3321 with a 330pf cap on input.
feed back resistors. from 22ohm to 1 K ohm
piggy backed R87 and R88 with 220pf capacitor.
RV1 full drive.

So results.
80M - 20W + ( blew my 2 Amp fuse)
40M - 18W - Seems stable.
30M - 17W BUT I get BAD feedback/ oscillation by just going PPT with no voice. The watt meter goes up to 18W!
20M - 15W  -  Same feedback issue
15M -  Same feedback issue.
10M -  Same feedback issue.

I have been thinking ans looking at other RD16 schematics, in alot of cases they dont use the feedback on the finals?
Also could this issue be cuase in the driver stage?

THanks guys, I am in over my head here....

73
10M - Same feedback issue

K9HZ <bill@...>
 

Ok just saw this.  Will do some testing here to better understand what works.  That 4:1 transformer is a good match for those parts judging by the power output. 


Dr. William J. Schmidt - K9HZ J68HZ 8P6HK ZF2HZ PJ4/K9HZ VP5/K9HZ PJ2/K9HZ

 

Owner - Operator

Big Signal Ranch – K9ZC

Staunton, Illinois

 

Owner – Operator

Villa Grand Piton - J68HZ

Soufriere, St. Lucia W.I.

Rent it: www.VillaGrandPiton.com


email:  bill@...

 


On Mar 11, 2018, at 6:57 PM, Nik VK4PLN <nickpullen@...> wrote:

Also,  RV1 Drive level has no effect on the feedback issues, happens just the same at 5W as at 15W...

K9HZ <bill@...>
 

Perfect Glenn.  Now if we can find current SMD parts to replace the 3904’s in place, this is a very nice upgrade. 


Dr. William J. Schmidt - K9HZ J68HZ 8P6HK ZF2HZ PJ4/K9HZ VP5/K9HZ PJ2/K9HZ

 

Owner - Operator

Big Signal Ranch – K9ZC

Staunton, Illinois

 

Owner – Operator

Villa Grand Piton - J68HZ

Soufriere, St. Lucia W.I.

Rent it: www.VillaGrandPiton.com


email:  bill@...

 


On Mar 11, 2018, at 11:02 PM, Glenn <glennp@...> wrote:

I completed my mockup of the PA driver stages.

as previous post:
  • used MPSH10's in all stages. 
  • Used FT37-43 toroids all wound with 10 turns 0.25mm wire either tri- or bi-filar according to original Sch.
  • Bypassed all emitter resistors with a 220pF cap.
  • swept from C80 input to T10 secondary (as wound)

Result is a very flat response. Better than 1db from 3Mhz to 54MHz.

glenn
vk3pe
<uBITX_PA_driver_stages_all emitter res bypassed with 220pF 120318.jpg>

Jerry Gaffke
 

All the *3904's in the driver stages are through-hole, the to92's have
a much higher max dissipation rating than the sot23's.
Jerry


On Mon, Mar 12, 2018 at 08:04 am, K9HZ wrote:
Perfect Glenn.  Now if we can find current SMD parts to replace the 3904’s in place, this is a very nice upgrade.