TVM Sound PK200

[Diabolical Artificer]

You can still get axial electrolytics, Vishay do some good ones (long leads). Just tried to look on Farnell, their site is down. They are easy to find using their filters.

Also F&T do axial's, Hi Fi collective or Watford valves may have them.

To check for hum use your meter on AC volts; anything over a few volts is suspect.

It would be worth going round taking some voltage readings after you've done to make sure all is well and to label your schematic for future reference of yourself and others who would find it useful in the future.

Andy.

[phykell]

Originally Posted by Diabolical Artificer

You can still get axial electrolytics, Vishay do some good ones (long leads). Just tried to look on Farnell, their site is down. They are easy to find using their filters.

Also F&T do axial's, Hi Fi collective or Watford valves may have them.

I think an axial, if I can find one, would be too large to mount under the chassis. The radial type is the one I need really - I just want to avoid the complication of trying to reduce the diameter though one of them does sit *very* close to the windings of the mains transformer, which doesn't seem too sensible to me.

To check for hum use your meter on AC volts; anything over a few volts is suspect.

The scope is on the speaker output - the amplitude increases as I increase the volume with no input signal so I do suspect the filter capacitors at the moment.

It would be worth going round taking some voltage readings after you've done to make sure all is well and to label your schematic for future reference of yourself and others who would find it useful in the future.

I'll do that - I really must get that schematic done

Oops, it's actually 600uF at 300V that I need. The only 40mm ones I can find are at RS - Wurth Elektronik at £14.00 for a 680uF and £7.70 for an 820uF. On that basis, I'm hoping it should be OK to use the 820uF part to replace the existing 600uF...

[Omegaman]

The 820uF ones will be fine.
Snap in types usually have enough 'tag' to make sound connections to.

[phykell]

Can anyone see a problem substituting a pair of 6550s for the KT88s? I've read around the subject and it seems that it's just a case of biasing them correctly which I'd have to do with new KT88s anyway.

I happen to have picked up another PK200 without any valves but the case is labelled as taking KT88s. As far as I can tell, there's no difference from the PK200 I have that's already running 6550As.

The price difference is substantial:

Pair of new, matched 6550s: £47.50 including delivery.
Pair of new, matched KT88s: £72.00 including delivery.

[Omegaman]

My mate uses a couple of PK200s as part of his reggae sound system and we use 6550s in those to keep costs down. No problem.
Just set the bias accordingly. His are set for an idle current of 28mA across a 1 Ohm cathode resistor on each valve.

[Herald1360]

That's biassed well back. The GE original data suggests around twice that, depending on Va and Pout! Should have a relatively easy time of it..... or is it just necessary with modern production bottles?

[phykell]

Originally Posted by Omegaman

My mate uses a couple of PK200s as part of his reggae sound system and we use 6550s in those to keep costs down. No problem.
Just set the bias accordingly. His are set for an idle current of 28mA across a 1 Ohm cathode resistor on each valve.

Thanks - the PK200 does have pots that can be easily adjusted but there are no test points. So to recreate his measurement method, I should connect a multimeter to pin 8 and then to ground via a 1 ohm resistor?

One of my units is marked up as -70V by each KT88 socket so I guess this just needs to be measured on pin 5 and adjusted accordingly or is measuring the current a better option?

[Omegaman]

Hi,
The -70v was a factory setting used before everybody became recently obsessed with bias settings. They would have been set with the voltage measured at pin 5 and the valves plonked in and off they went.
It would be kinder to the valves to adjust it a bit more accurately.
Pin 8 of each valve needs to be connected to ground with a 1 Ohm resistor instead of a direct wire connection. Connect your meter black probe to ground and measure the voltage at pin 8. A voltage of 28mV equates to 28mA cathode current.
Ideally, the anode voltage needs to be measured and the required idle current worked out from that and the anode dissipation of the valve etc. An easy way is to use the online 'Weber Bias Calculator'.

[snowman_al]

Hi Omegaman,
I possibly have the same itch as Herald1360, why such a low bias current?

KT88/6550s are rated at 40+ watts dissipation.
I assume the amp has 500 volts HT so at 28mA each valve is dissipating just 14 watts.

phykell,
Out of interest what voltage do you measure on both pin 5s of your working example and did you alter anything?

Alan

[phykell]

Originally Posted by Omegaman

Hi,
The -70v was a factory setting used before everybody became recently obsessed with bias settings. They would have been set with the voltage measured at pin 5 and the valves plonked in and off they went.
It would be kinder to the valves to adjust it a bit more accurately.
Pin 8 of each valve needs to be connected to ground with a 1 Ohm resistor instead of a direct wire connection. Connect your meter black probe to ground and measure the voltage at pin 8. A voltage of 28mV equates to 28mA cathode current.
Ideally, the anode voltage needs to be measured and the required idle current worked out from that and the anode dissipation of the valve etc. An easy way is to use the online 'Weber Bias Calculator'.

Thanks - I've watched YouTube videos and read around the subject but that's just wrapped it up for me. I'm guessing that if I went for a matched pair of KT88s, it would likely be fine but I'm sure the cheaper 6550s are the way to go.

[phykell]

Originally Posted by snowman_al

Out of interest what voltage do you measure on both pin 5s of your working example and did you alter anything?

Sorry about the delay - I'll get back to you when I get the chance to have a go. I still need to complete and submit my (almost complete) schematic for these units as well.

[phykell]

With the power valves removed, I got the following measurements, grounding to the chassis:

Power valve closest to the MAINS TRANSFORMER
Pin 3 : 559v
Pin 5 : -82.9
Pin 8 : All over the place

Power valve closest to the OUTPUT TRANSFORMER
Pin 3 : 544v
Pin 5 : -82.8
Pin 8 : All over the place

I don't have a 1R resistor but will try with a 10R and adjust the calculation as necessary - do the power valves have to be fitted?

The control grid seems a bit high to me at -82.9v - should I expect distortion or other unfortunate side-effects?

[phykell]

Originally Posted by snowman_al

Out of interest what voltage do you measure on both pin 5s of your working example and did you alter anything?

Hi Alan,

As mentioned, pin 5 is at around -83v. I haven’t modified the amplifier at all though I did replace just about all of the capacitors and all of the electrolytics. I’m hoping that’s why the voltage is now so high but I intend to drop it to -70v as a starting point and then look at adjusting it properly.

I’ll take some photos of the other two I have and post the same plus my progress (hopefully) fixing them

[snowman_al]

Originally Posted by phykell

With the power valves removed, I got the following measurements, grounding to the chassis:
...
I don't have a 1R resistor but will try with a 10R and adjust the calculation as necessary - do the power valves have to be fitted?

The control grid seems a bit high to me at -82.9v - should I expect distortion or other unfortunate side-effects?

No, no undue side effects, maybe just a little less power output. The closer you get to 0 volts the more current you dissipate through the valve.

Yes the valves have to be there or you do not pass any current through the 10 ohm resistor to measure the voltage ... (that will be 0.28 volts going with Orangeman's recommendation.)
In fact be careful pulling the output vales out at all, the HT voltage will go up and might exceed your capacitor maximum voltage rating.

As you say, put the vales in and set their pin 5s (grid) volts to -70 volts to start with, until you do the cathode resistor mod.
Also Make Sure you have a load (speakers or dummy load) connected when setting / testing.

Alan

[phykell]

Here's are some pictures of the other two PK200s I have:

https://1drv.ms/u/s!AkB_A0SO1G6dlQYYLI2tHTqe9OKB

https://1drv.ms/u/s!AkB_A0SO1G6dlH2C7gnDIEJ4J8fR

They're a bit grotty right now but they started to clean up really well. Note the different transformers on one of them though - I'm told that the other two are Partridge transformers but I've no idea what these other ones are.

I'm not sure about the 6550s as an alternative to the KT88s. The 6550s that I looked at are supposed to only have an anode voltage of 660v instead of 800v on the KT88s...

Underneath views:

https://1drv.ms/u/s!AkB_A0SO1G6dlHjrhauzQYEVvWYw

https://1drv.ms/u/s!AkB_A0SO1G6dlQnSsorn11tbEkQk

One of them has some blown filter capacitors if you look carefully:

https://1drv.ms/u/s!AkB_A0SO1G6dlHllT3gn4AZJ9Y26

[snowman_al]

The transformers might all be 'Partridge'.
The older chassis (non IEC mains connector) with the C core ones are typical Parmeko (Partridge & Mee). But they come in lots of ordinary styles too.

I notice you changed the 'mustard' coloured capacitors on the first one. Keep the originals, they are almost certainly perfect.

You only have 550 volts HT, so why worry about using 6550s? They will be fine.

Alan

[phykell]

Originally Posted by snowman_al

The transformers might all be 'Partridge'.
The older chassis (non IEC mains connector) with the C core ones are typical Parmeko (Partridge & Mee). But they come in lots of ordinary styles too.

I notice you changed the 'mustard' coloured capacitors on the first one. Keep the originals, they are almost certainly perfect.

You only have 550 volts HT, so why worry about using 6550s? They will be fine.

Alan

Thanks - you pointed out what should have been obvious to me (the voltage) and I’ve now ordered the 6550s

Thanks for the info on the transformers as well - I haven’t seen ones like this before but they are at least more interesting to look at than the more conventional ones. I wonder if the newer ones actually managed to improve on them in any way.

On the caps, I just replaced them as a matter of course but I did at least keep the old ones. I didn’t know that the mustard ones would be of any use, assuming that they were so old anyway, they may be breaking down under use or their values may have changed over time.

[Diabolical Artificer]

If your leaving the 1 or 10 ohm sense resistor in place try and use a 5w rated resistor, 3w at least. Welwyn WW2 resistors are ideal here, they're rock solid though cost a bit more. We don't want the resistor going high in value under fault condition's so it's best to be safe than sorry by fitting an over rated decent part.

Once you've set bias, if you connect your DMM across the OP stage, EG one lead to pin 5, tother lead to pin 5 you can balance the OP by gently turning one pot CW, the other CCW, and vise versa. You are trying to "null" the reading, IE get it as near to 0v as poss. You'll see it will vary from 0.054v, to -0.054v for instance. If you can get it either + or - 0.008v say it will be in good balance meaning THD will be reduced and CMRR improved (less hum + noise). This will be fiddly to do but worth it. Try giving the pots a squirt of contact cleaner before doing so.

Those transformers are the ducks guts, built like the proverbial brick outhouse. Partridge tfmr's as you probably know are well sort after, often worth more than the amplifier itself. The amps though rough are a nice find.

Andy.

[Herald1360]

Originally Posted by phykell

.


On the caps, I just replaced them as a matter of course but I did at least keep the old ones. I didn’t know that the mustard ones would be of any use, assuming that they were so old anyway, they may be breaking down under use or their values may have changed over time.

Just Google "Mullard mustards" ! In some circles they have mythical status....

They're simply early plastic film (rather than paper) dielectric and as such their longevity is much better.

You may well be able to sell the old ones for more than you paid for their replacements