Rogers HG88 III Hum

[PaulR]

I recently dug out the Rogers amp that was the subject of a discussion on hum a few years ago following several years of disuse. I had already replaced all the caps in the amplifier and any out of spec resistors. I had also replaced the caps in the power supply but wasn't happy with the job I had done so I have just obtained new ones and fitted then using plastic securing clips which I had omitted last time. I also replaced the rectifier diodes.

Anyway, it sounds good if a little excessive in the treble range to my ears but there is an annoying hum in one channel independent of the setting of the volume control. I have tried swapping the output valves round in various permutations but that does not help. The other channel is completely silent.

Before I return it to its box in the loft can anyone suggest anything I can try to eliminate the hum. I have read about problems with the output transformers in Rogers valve amps, but except for the hum both channels sound exactly the same.

Thanks

[ukcol]

Circuit here:-

http://www.valve-radio.co.uk/literat...-instructions/

Push-Pull output stages often rely on the output circuit being in balance in order to cancel out any residual ripple on the main H.T. supply, so that may be worth checking.

If there is no problem there and you have an oscilloscope, you could check the ripple voltage across across each H.T. smoothing/decoupling capacitor and compare each reading with the good channel.

Unusual voltage doubler power supply BTW.

[Bookman]

Does the same balance issue affect DC offset which I know will impact upon txfr losses but would the offset component be reflected/magnified on the txfr secondary? More importantly, how would one effectively isolate it?
Just curious!

[PaulR]

Thanks Colin.

Could you just explain how I can check that the output circuits are in balance, please?

I do have an oscilloscope by the way

Thank you

[ukcol]

If you measure the voltage at the cathodes of the outputs valves (pin 7) with PRV2 in the mid position the voltages should be about the same. A small difference in the cathode voltages will not be significant.

Note, when you are done put PRV2 back to its original position (assuming the original output vales are in their original positions). There does not appear to be any instructions on setting up PRV2 in the document I provided a link to, but I suspect it would be done to obtain a symmetrical sine wave at the output. That is that it is probable there to balance the gains of the output valves rather than to provide balance of DC conditions.

Some high quality output transformers are wound using a method that gives equal DC resistances for each half of the primary. If the resistances of these halves are equal in your transformers you can check for DC balance by comparing the DC voltage drop across each primary half.

Make your measurements with the volume control turned down to zero.

[Diabolical Artificer]

Might be a htr cathode short on one valve. I'd power it up sans valves and scope the HT. If that's clean pop the valves back one at a time checking the OP of each on a scope.

Andy.

[ukcol]

Originally Posted by PaulR

..........there is an annoying hum in one channel independent of the setting of the volume control. I have tried swapping the output valves round in various permutations but that does not help.

I had dismissed the possibility of a h/k fault Andy because of Paul's report as above.

[PaulR]

Thanks for the responses.

I measured the cathode voltages and they were similar on the bad side (0.363 and 0.294). They were in fact further out on the good side!

I then started to look at the anode connections to measure the resistance of each side of the primaries and discovered a bad solder joint on one of them. Re-making this has cured the problem. It would also explain a phenomenon I had dismissed as imagination in that I sometimes thought that the fault had briefly cured itself.

Anyway, thank you for the help. I will now bring it into full use, albeit with the treble control turned well down!

[ukcol]

Well done Paul, that's good news.

[hansomcommon]

Superb amplifiers, please use it! Some of the best unforced base that I have heard from a "modern" valve amplifier.

[ukcol]

Originally Posted by Bookman

Does the same balance issue affect DC offset which I know will impact upon txfr losses but would the offset component be reflected/magnified on the txfr secondary? More importantly, how would one effectively isolate it?
Just curious!

I'm not sure exactly what you are asking, however, in a single ended valve amplifier the DC current flowing in the primary magnetises the core and the transformer has to be made bigger to avoid saturation. In a push pull amplifier the magnetizing effect of the DC currents in the output valves cancel allowing the transformer to be smaller. In Hi-Fi push pull valve amplifiers the output transformers tend to be large because high inductance is required to give a good low frequency response and not because of the risk of saturation.

Transformer design is a very complicated subject and the little I know has come from reading in this forum and elsewhere. A number of our forum members are transformers experts and will no doubt give you a better and fuller answer.

[PaulR]

As a matter of interest are they known to be a bit "heavy" on the treble? It could be the speakers I am using. They are B&W DM4s which have two tweeters.

[PJL]

I would think it's the speakers. I got a pair recently for very little and I find them a bit harsh and my hearing falls away at about 10K. If they haven't been used for a while I guess they may improve with use. I took a look at the crossovers but they appear to be solid dielectric and not electrolytics so didn't bother replacing them.

[PaulR]

Speakers are so subjective it is difficult to know what to buy.

I once bought a pair of Ditton 44s but SWMBO took one look at them and said "NO" so they spent some time in the loft and I eventually sold them to a friend.

[Ted Kendall]

I wouldn't be too quick to condemn the DM4s - in good condition they are well-balanced and very competent. The remark about two tweeters is misleading - in fact it has a tweeter and a super-tweeter - a question of extension rather than spectral balance.

[PJL]

This was using a solid state Roksan Caspian amp which would bear little resemblance to the spec of an amp from the early 70's. I haven't tried them on a valve setup yet.

[PaulR]

Following the success of removing most of the hum as described in this thread, there was some residual hum which I managed to cure by swapping the output valves around. These must be far from matched as they are different makes and ages and I was wondering whether it would be possible to match them better by using the variable resistors in the cathode circuits.

I am not sure how to go about this. Should I try to balance the DC on the anodes? The two halves of the output transformer primaries do not have the same resistance so I am not sure whether this is possible or desirable.

I have a decent, if older, HP oscilloscope which I could use to test the signal output voltage on each anode if that would be the way forward but, again, I am not sure whether this is desirable in view of the different resistances of the halves of the primaries.

Can anybody suggest what I can try?

Thanks

Paul

[Diabolical Artificer]

As your amp OP stage uses a shared cathode resistor for the two ECL86 pentodes, there's not a lot you can do, unless you take this out and put two separate cathode resistors in it's place.

As regards the different OPT primary halves, it shouldn't make much difference as long as it's impedance is roughly the same. Have you scoped the OP with a sinewave IP and checked to see if the negative and positive going peaks are the same amplitude? You can check balance with a decent DMM too.

Obviously the best way to insure matched OP of the PP stage is to use matched ECL86's, both triode and pentodes, but that will cost a fortune if you can get a good few to test. You'd have to buy about 40 I reckon to find a matched pair.

Andy.

[PaulR]

Thanks Andy

The cathodes on the MkIII share a sort of potentiometer being connected to each end of the track and an earthed slider so there is a theoretical means of adjustment.

I will check the amplitudes of the outputs using the 'scope and see whether I can balance them.

Thanks

[PaulR]

The RMS voltages on the anodes of the two halves of each channel were about 1 to 1.5v different depending on the frequency with the average actual voltage at around 6v. I used 0.1uf caps to isolate the probes. Changing the valves round or moving the sliders on the cathode resistors made no difference so I have put them back to the central position.