~250kHz pings from neve 1073 type line output stage

[HenryL]

With a scope on the line out of a neve 1073 copy (ie. I'm listening on the line out side of the output transformer) I'm seeing high frequency pings occurring. It looks like they are occurring fairly regularly at about every 5ms, though I'm not currently quite sure how regular they are - they keep coming though, it's not just the odd one. There's a spike then the oscillation decays away. The frequency of the ringing is approx 250kHz.

It only occurs in the final TX output driving circuit stage (around neve BA283 TR1/2/3). If I ground the input to that stage it carries on pinging. I don't see any sign of it on the power supply.

I put the lid back on the unit to ensure it was normally screened but it made no difference.

The initial spike can be ~200mV.

Does this sound familiar to anyone?

 

[RoadrunnerOZ]

Sounds like a failing capacitor or transistor. Does it still occur if you disconnect the transformer? 250KHz ring occurring at an approximately 200Hz frequency is a bit strange - I’d scope the electrolytics on the board - especially those involved with supply filtering - and any associated hanging off that board.

 

[HenryL]

Sounds like a failing capacitor or transistor. Does it still occur if you disconnect the transformer? 250KHz ring occurring at an approximately 200Hz frequency is a bit strange - I’d scope the electrolytics on the board - especially those involved with supply filtering - and any associated hanging off that board.

Disconnecting the OP transformer on a 'BA283' - that'll effectively just kill the whole OP stage won't it ?!

I didn't see these pulses on the supply line when monitoring it at the last stage (at the +supply at the OP TX). I was slightly surprised not to see anything at all there, but I'm taking that to rule out the whole supply/filtering caps.

I posted the question straight away in the faint hope that someone might just say 'ah that old problem, definitely a failing 2N3055' or something !
I need to do some more systematic probing.

Must buy you a beer.

 

[RoadrunnerOZ]

Disconnecting the OP transformer on a 'BA283' - that'll effectively just kill the whole OP stage won't it ?!

Yes - you will need to insert some form of dummy load - the idea is to isolate whether the transformer is part of this weird ringing. Is the time between pulses of 250KHz of 5ms constant or drifting?

 

[RoadrunnerOZ]

Also is there any component on the board getting hot - do you have freezer spray to systematically freeze components to see if the fault disappears?

 

[RoadrunnerOZ]

There are other caps normally attached to the 283 board but I’m not sure exactly what is ancillary to this board in your build. Do you have a full schematic for your preamp?

 

[RoadrunnerOZ]

The symptoms are indicative of a capacitor charging up then breaking down, shorting the supply, recharging, breaking down, shorting the supply, recharging….. in an endless cycle.
But it can also be a semiconductor like a transistor or a protection diode breaking down. It’s pretty easy to swap out electrolytics or transistors - all cost only a few cents anyway except the 2N3055 which only cost a few dollars.

 

[HenryL]

Yes - you will need to insert some form of dummy load - the idea is to isolate whether the transformer is part of this weird ringing. Is the time between pulses of 250KHz of 5ms constant or drifting?

OK - see what you mean. What sort of dummy load (assuming you mean a resistive load) would you suggest in this case ? Low R to let it idle near the normal point vs higher R to give it some signal load but screw up the voltages?

 

[HenryL]

I have traced the audio circuit to verify that it is the same as the Neve schematic, there are just some slight variations in the sensitivity selector values, feedback and gain controlling resistors, filter caps etc. - the main addition/ancillary is a basic 4-LED output level monitor circuit (which I haven't traced out) based around an LM399 quad comparator and TL072 op amp chip. The input to the metering is taken from Neve BA283 point 'F' unbalanced output (or 1073 point 'P'). Amongst other things I'll try checking at 'point F', cutting the connection to the metering circuit in case it is coming back into the audio circuit from there, though I think that's less likely.

The 2N3055 isn't getting hot but I haven't checked the rest yet. No freezer spray but about time I got some.

Other than that - time to roll up shirt sleeves and wade in again...

(I'm not completely sure how regular the pulses are - I'm going to test to capture that more clearly)

cheers

 

[RoadrunnerOZ]

If the Zener in the comparator breaks down completely the metering wouldn’t work - does the metering work? Are there any electrolytics hanging off the 283 board? This is like a charge/fail/discharge (capacitor) or a heat/fail/cool (semiconductor junction failure or even cracked resistor) problem. Question; does it happen immediately on turn on or take time?

 

[HenryL]

If the Zener in the comparator breaks down completely the metering wouldn’t work - does the metering work? Are there any electrolytics hanging off the 283 board? This is like a charge/fail/discharge (capacitor) or a heat/fail/cool (semiconductor junction failure or even cracked resistor) problem. Question; does it happen immediately on turn on or take time?

The metering seemed to be working normally though I wasn't paying close attention to it whilst fiddling with the TX zobel, which is when I first noticed the spiking. I'll be watching out for anything anomalous whilst debugging this spike behaviour more systematically, and to see if it starts immediately on turn on etc.

If it had been an actual Neve 283 board then I guess I'd say: yes, it has 100uF in series with 1K5 between terminals J (gnd) and K, as per 1073, though it's actually all on the one board in this case. There are 3 other electrolytics in 'interesting' parts of the final drive circuit to look at as I see it.

 

[RoadrunnerOZ]

Check ‘em all- if there’s spiking on the top of a cap and it gets less on the other side of a component attached to it the it’s likely the culprit. Mind you sometimes it’s just quicker to replace than trace.

 

[jokeramik]

What of type of power supply do you use? Linear regulated or switch mode?
Best regards!

 

[HenryL]

What of type of power supply do you use? Linear regulated or switch mode?
Best regards!

It's a linear supply. A line-lump transformer with the rest of the PSU, regulators etc. inside the rack case.

(I didn't see any of these spikes/pings on the power supply at the output stage, but I'm going to go back and check everything again with a little more time)

Cheers

 

[RoadrunnerOZ]

There are 3 other electrolytics in 'interesting' parts of the final drive circuit to look at as I see it

Are these equating to M and N on the original 1073?

 

[jokeramik]

Another problem could be the insufficient decoupling of the supply voltages of the individual amplifiers. This is how this phenomenon occurs. Sometimes also to so-called motor boating.
Best regards!

 

[RoadrunnerOZ]

Failing capacitor(s) on supply rails can cause this problem - but in this case the main supply rails would appear clean, however there are other areas of decoupling in this circuit where it can occur - that’s why all electrolytics should be checked or just do a recap. There aren’t that many to replace. I recently did a complete audio coupling and supply rail filter recap of a 36 channel Neve console plus replaced all the old Kingshill power supplies with new medical grade power supplies. The desk is now dead quiet.

 

[HenryL]

Are these equating to M and N on the original 1073?

No I meant onboard labelled C4,6,7 on the 'BA283 AM&AV' (EX/10283) Neve schematic. (OK...C6 does connect to the M +ve supply line but not externally like the C7 and C8 filter caps on the 1073 diagram). C4 is in a feedback circuit, C6 is bypassing the 2N3055 emitter biasing/balancing resistors to the +ve supply rail. C7 is the unbalanced output at F (which then goes to the metering circuit in this case) as I read it.

I'll try to strike a balance between investigation and blanket replacements

 

[HenryL]

Another problem could be the insufficient decoupling of the supply voltages of the individual amplifiers. This is how this phenomenon occurs. Sometimes also to so-called motor boating.
Best regards!

Hi - as Roadrunner OZ said, the bad behaviour doesn't appear on the supply line (though I only had time to check it rather quickly and will double check, today hopefully. It is certainly not doing what I would call motorboating, which is a low frequency 'chugging' oscillation (which I've come across a couple of times in guitar amps). It's a repeating pattern of a 'long' silence then a ping which rings at a very high frequency.
cheers

 

[RoadrunnerOZ]

When you say long silence I think you originally said 5mSec or is it 5Sec?
How old is this unit?