2445B vertical non-linearity other issues.


Robin Szemeti
 

I have a 2445B that has gone a bit odd recently.

Ch2 and to a lesser extent CH1 have developed vertical non-linearity. Put a sine wave through it and it is obvious, a sort ot cross-over distortion around the centre of the tube .. Moving the trace around with the vertical position control it appears the distortion stays stationary and the trace moves through it. I also noticed the text had moved vertically off the bottom of the display ... and if I move CH1 or CH2 beams into the upper part of the display, the text starts to jitter and shake .. this does not happen with CH3 or 4

This non linearity does not affect CH3 or 4 so I think the vertical amp U600 is fine, and swapping with a good spare I have did not improve it. I'm beginning to suspect the channel switcher IC, U400 ... do the symtoms sound typical?


Roger Evans
 

Vertical nonlinearity on Ch1 and Ch2 but not on Ch3 and Ch4 is typical of a fault in U400. The other symptoms are new to me but I am sure someone here will offer some insight. It would be worth checking the resistors that set the emitter currents for U400, I need a large screen in order to look up part numbers!

Regards

Roger


Robin Szemeti
 

I'll try the emitter resistors, but there are so many cases of U400 failure that I suspect it is going to be that. Which sucks. A lot.


Harvey White
 

In a totem pole pair (NPN on the top, PNP on the bottom), if no current flows through the transistors when at zero voltage in, then there's going to be a dead spot until the input voltage rises enough to turn on the NPN, or lowers enough to turn on the PNP.

This is generally fixed by allowing a certain fixed current to go through both NPN and PNP so that the NPN turns on while the PNP is still on.  It increases the power consumption, but minimizes what's called "crossover distortion".

The suggestion to check the emitter currents is precisely this for the output stage.  The physical design of the output stage can vary (the example I used is generally used in audio amplifiers, NPN transistor as an emitter follower for + voltages, PNP transistor ditto, but for negative voltages).

For the typical Tektronix scope, you have an emitter coupled differential pair (likely towards the beginning of the amplifier, and then a DC coupled amplifier from then on.  The turn on/turn off points of the transistors is critical to the crossover distortion issue.  That's the quiescent emitter currents of the amplifiers.

Harvey

On 10/4/2021 4:51 PM, Robin Szemeti wrote:
I'll try the emitter resistors, but there are so many cases of U400 failure that I suspect it is going to be that. Which sucks. A lot.





Robin Szemeti
 

Yep, I am familiar with class A/B amp biasing .. I have built "a few" audio amplifiers.

There is only one resistor, R401 I Bias, that seems relevant. From other threads it looks like the fault with this device usually lies on the thick film itself. It seems in the past, there was a re-bake programme for these devices which recovered 70% of them ... but I'll check R401 and see what gives ...


Harvey White
 

Part of the reply is to help you, if needed, and the other part is to help people who may want a take on the problem itself when they search the topic.   If I get it wrong, then the corrections are useful.

if it's the hybrid, the rebake program is problematical, from what I remember.  I don't have a 2465 of any variety, but I do have a 2430A as a second/third/etc scope.  Most of what  I do works well enough with workable 7904's and the occasional 7103/4. My portable scopes (468, 2430A, and a digital japanese scope) top out at about 150 Mhz.

Good luck.  As I understand it, be careful when replacing the hybrid.

Harvey

On 10/4/2021 8:11 PM, Robin Szemeti wrote:
Yep, I am familiar with class A/B amp biasing .. I have built "a few" audio amplifiers.

There is only one resistor, R401 I Bias, that seems relevant. From other threads it looks like the fault with this device usually lies on the thick film itself. It seems in the past, there was a re-bake programme for these devices which recovered 70% of them ... but I'll check R401 and see what gives ...





Roger Evans
 

Given the cost of a replacement U400 and the oddities you observe on the readout, it is well worth checking the waveforms at <51> and <52> in the service manual schematic <6>, this is with the readout turned off. They should be closely balanced about ground and the transistor array U475 is a likely culprit for any problems here.

Best of luck,

Roger


Robin Szemeti
 

I'm building a replacement U400 from discrete surface mount components. I'll let you know how it works out.