Faulty A Designs EMPEQ - deceasing volume

[Phantasm]

I have an A Designs EMPEQ 500 series equalizer that has just gone faulty in that when the EQ is engaged the volume slowly decreases until barely audible - taking maybe 20 seconds to get to that state. Anyone ever heard of a fault like this? I emailed A Designs and they say they've never heard of this happening.
The unit is 2nd hand so unfortunately no warranty or anything but has been working fine in my rack for just under a year. A Designs suggested trying another rack but don't know anyone local to me that uses 500 series and all my other modules are fine so can't be that plus checked and have the requisite 48V, +16V and -16V so don't think its the rack. Popped the cover and can't see anything un-towards. My rack is an IGS Audio panzer and has a good beefy supply at 420mA per slot plus has only 5 slots filled out of 10 at the moment so can't be lack of power.
Any ideas anyone? Thanks in advance for any tips.

Justin,

 

[JohnRoberts]

Is it overheating? while that does not generally degrade gracefully.

JR

 

[Ike Zimbel]

What happens if you fiddle with the Eq switch? I don't know the device at all, but if the Eq In/Out switch was a push-push type, maybe the spring is weak and it's disengaging? This is a stretch, for sure. Pictures, inside and out, would help.

 

[TwentyTrees]

Do your other modules in the rack work OK while you're experiencing the fault you describe? If so, unlikely to be the main rack power supply (though per-slot filtering or regulation may be worth looking at, if you're looking at the rack).

Otherwise, heat does sound like the kind of mechanism that could cause a gradual onset fault like this. From a quick Google, it seems the EMPEQ 500 is a Pultec-style EQ. I've not seen any internal pics, but if it follows the solid state Pultec topology then there would only be a few active devices to investigate. Pics would certainly help.

Is the level decrease consistent? Does it always fall by a certain number of dB? Is it uniform across the frequency spectrum?

 

[Phantasm]

Is it overheating? while that does not generally degrade gracefully.

JR

Hi John, no doesn't seem to be overheating. I plugged into an extender card and left powered on eq engaged for a while last night but didn't get hot at all.

What happens if you fiddle with the Eq switch? I don't know the device at all, but if the Eq In/Out switch was a push-push type, maybe the spring is weak and it's disengaging? This is a stretch, for sure. Pictures, inside and out, would help.

It's a toggle switch and seems in good working order so doesn't look like the switch. I can take some pics to upload if it helps.

Do your other modules in the rack work OK while you're experiencing the fault you describe? If so, unlikely to be the main rack power supply (though per-slot filtering or regulation may be worth looking at, if you're looking at the rack).

Otherwise, heat does sound like the kind of mechanism that could cause a gradual onset fault like this. From a quick Google, it seems the EMPEQ 500 is a Pultec-style EQ. I've not seen any internal pics, but if it follows the solid state Pultec topology then there would only be a few active devices to investigate. Pics would certainly help.

Is the level decrease consistent? Does it always fall by a certain number of dB? Is it uniform across the frequency spectrum?

Yes have 5 other modules that all work fine. I tried in another slot and did the same thing. The best test would be in another rack but don't have another one and don't know anyone else local with 500 rack but as you say it seems unlikely to be the rack itself.

Yes it's a Pultec clone - really nice EQ but seems not to be popular with Pultec purists. People always seem to compare it to the original version which it is obviously not going to compare to. The new Pultec 500 series would be the closest comparison but never seen a shootout between the 2.

Yes seems to take a similar amount of time to drop each time plus appears to be losing a few dB every 5 seconds or so - not timed it exactly but rough estimate. and seems to happen equally on different settings so I would say yes uniform across the frequency spectrum. I'll take some pics of the guts and post here.

Thanks for the assistance everyone!

J.

 

[TwentyTrees]

Follow-up questions - does it lose the same amount of dB each time, and if so how many? Do all the EQ controls work as normal while the unit is experiencing the fault condition?

Assuming that the topology is the same as the Pultec, then the answers to these questions may potentially help us narrow down the issue to the filter section or the make up gain amp.

 

[Phantasm]

Follow-up questions - does it lose the same amount of dB each time, and if so how many? Do all the EQ controls work as normal while the unit is experiencing the fault condition?

Assuming that the topology is the same as the Pultec, then the answers to these questions may potentially help us narrow down the issue to the filter section or the make up gain amp.

I didn't try adjusting the controls while it was happening but I will give it a go later today when home from work and report back plus will try to accurately measure the rate of volume reduction and dB level loss. I'll also take some pics and post.

Thanks for the suggestions - very much appreciated.

J.

 

[Ike Zimbel]

If it's a fairly faithful reproduction of a Pultec, it's a passive Eq with an amplifier. So, it sounds like the amplifier is the issue (meaning you would still hear the Eq'd signal, but at a low level). If it has tubes in it, I would suspect that the whatever DC/DC conversion circuitry they're using for the B+ has an issue.

 

[Phantasm]

Follow-up questions - does it lose the same amount of dB each time, and if so how many? Do all the EQ controls work as normal while the unit is experiencing the fault condition?

Assuming that the topology is the same as the Pultec, then the answers to these questions may potentially help us narrow down the issue to the filter section or the make up gain amp.

Did some more tests and with my signal initially at about -2dB it drops to about -25dB but doesn't seem to drop any lower or cut out. It seems that the eq controls do still function while losing volume but then while playing about I found that at certain points the volume snaps back to full and I noticed that certain frequency settings appear to be quite happy. This suggests to me that a particular part of the circuit that deals with a certain set of frequencies - I think maybe 6K and above and appears mainly when using extreme cut and boost settings of the high frequency and am sure that adjusting the bandwidth between broad and narrow also seems to be involved. The attenuator select didn't seem to affect the results so maybe narrows it down to the high frequency selection. I'm going to reply to A Designs with these findings also in the hope that he has a suggestion or see if he's willing to give me the schematic.

Attached pics of the PCB and driver module pcb removed from the main pcb although from findings above I doubt that it's the driver anyway but added since I took the pics before retesting.

 

[Phantasm]

If it's a fairly faithful reproduction of a Pultec, it's a passive Eq with an amplifier. So, it sounds like the amplifier is the issue (meaning you would still hear the Eq'd signal, but at a low level). If it has tubes in it, I would suspect that the whatever DC/DC conversion circuitry they're using for the B+ has an issue.

Hi Zimbel,
Yes it's reviews suggested that in terms of the circuit it was a faithful recreation but minus the tubes in the amplifier therefore giving it a more modern sound than a real deal pultec.

As you can see above I think I've narrowed it down to the eq circuit.

J.

 

[JohnRoberts]

Have you tried percussive troubleshooting ( like tapping components with a pencil eraser)?

Then there is spray cold... etc.

JR

 

[pucho812]

Have you tried measuring power rails in the unit while outside of chassis on an extender?
See if they drop in the unit? See how that are at the ad-10.
Outside of that I would bet a fault in the ad-10 as that is the makeup amp. Looks very quad 8 style. So you might get away with an am10 schematic to help trouble shoot

 

[Phantasm]

Have you tried percussive troubleshooting ( like tapping components with a pencil eraser)?

Then there is spray cold... etc.

JR

Haha! percussive trouble shooting - never heard it described like that before! Not tried as been trying to isolate the frequencies that are causing the problem.

 

[Phantasm]

Have you tried measuring power rails in the unit while outside of chassis on an extender?
See if they drop in the unit? See how that are at the ad-10.
Outside of that I would bet a fault in the ad-10 as that is the makeup amp. Looks very quad 8 style. So you might get away with an am10 schematic to help trouble shoot

Please excuse my ignorance but what are ad-10 and am10 referring to?
I measured the power rails via an extender card and all are nice and healthy plus 5 other modules are completely happy so think it's looking like the EQ part of the empeq circuit. What make you think it's the make up amp rather than the eq out of interest? Did you read my post 5 above? describing the behavior while adjusting the eq?

 

[pucho812]

Please excuse my ignorance but what are ad-10 and am10 referring to?
I measured the power rails via an extender card and all are nice and healthy plus 5 other modules are completely happy so think it's looking like the EQ part of the empeq circuit. What make you think it's the make up amp rather than the eq out of interest? Did you read my post 5 above? describing the behavior while adjusting the eq?

Am10 is a discrete opamp made by quad 8. Since a-designs borrows from the quad eight lineage the ad10 would be somewhat similar. In this case it is used as a makeup amp.
as you know the pultec design in most simplistic terms is a passive eq + make up amp. This covers insertion loss.
I would be willing to suspect a fault in the makeup amp over the passive section
Any day. But stranger things have happened. I once had a pultec where the owner insisted it sounded like 5k. His words. Turns out one of the wires broke off the lf boost pot. After 50 years in service as an eq, it’s bound to happen.
the result before fixing was a gentle bump of around 5k with a wide q. Once I fixed the wire, it was back to normal.
If your certain it’s the passive section check every connection and component.
However starting off at one volume level and slowly dropping down in volume leads me to thinking a fault in the makeup amp.

 

[Phantasm]

Am10 is a discrete opamp made by quad 8. Since a-designs borrows from the quad eight lineage the ad10 would be somewhat similar. In this case it is used as a makeup amp.
as you know the pultec design in most simplistic terms is a passive eq + make up amp. This covers insertion loss.
I would be willing to suspect a fault in the makeup amp over the passive section
Any day. But stranger things have happened. I once had a pultec where the owner insisted it sounded like 5k. His words. Turns out one of the wires broke off the lf boost pot. After 50 years in service as an eq, it’s bound to happen.
the result before fixing was a gentle bump of around 5k with a wide q. Once I fixed the wire, it was back to normal.
If your certain it’s the passive section check every connection and component.
However starting off at one volume level and slowly dropping down in volume leads me to thinking a fault in the makeup amp.

Ahh ok - thanks for the knowledge. I was unaware of the link to quad 8 and also I'm afraid not familiar myself other that knowing their reputation. You may be entirely correct, I'm going to send A Designs the info I now have to see what he suggests so if you don't mind I'll add your info as he obviously knows the Pultec inside out too. Thanks again.

J.

 

[Phantasm]

A little update - I went away on holiday for a week and then came home to a heat wave to find that the EMPEQ seemed to be working fine so was unable to recreate the fault. That is until using again and the fault has reappeared but this time definitely seeing it only when using 5K setting on the high band so will get looking again and report back. I'll start with disassembling from the front plate and back plate to check the switch wiring and soldering for any iffy looking joints.

 

[Phantasm]

Finally got time to strip the unit down and investigate the switches and solder joints and appear to have fixed the problem.

The switches all looked perfect - no loose wires. Solder joints looked good - most of the pcb appears to be machine soldered with some hand soldered parts also. Decided to re-solder the switches and headers for the panel mount switches and now appears to be functioning perfectly again after doing some thorough tests of each frequency. All I can think of is that 1 or 2 of the hand soldered joints had succumbed to its age and just needed revitalizing.

 

[TwentyTrees]

Great news! Glad you were able to bring it back to full function.

 

[Phantasm]

Great news! Glad you were able to bring it back to full function.

Thanks - glad to have it working again but also glad to be able to feedback on here in case a similar problem appears with anyone else which could apply to other similar eq devices.