switching thump...

[wintermute]

OK So I finally got around to putting a direct input on the back of my Integrated Series 200 Playmaster amp. Initially this is so I can run my HT PC direct into the amp bypassing the preamp section, but still have the rest of the preamp switching available for other inputs.

Basically I stuck a DPDT switch in between the 1K output resistors of the preamp and the 1uf input caps on the power amp. The switch is a simple toggle switch, center on off on.

Switching to the off position presents no problem at all. However switching to either preamp, or direct input results in a large thump from the speakers. The amp has relays on the speaker output and I could put a three pole double throw switch in and simply switch the relays off when switching inputs, which would probably do the trick, but I was wondering if there was anything else I could do.

I'm a little concerned that when I make a new preamp (as the excercise has definitely highligted there are deficiencies in the built in one which I had long suspected) that depending on the design I may still have a problem. The amp is eventually going to become my sub amp, and would be connected to a pre with an active crossover, and probably a B1 buffer on the outputs. I suspect I won't have a problem, but since I don't know exactly what causes the thump I'd like to know what the usual cause is and what workarounds if any there are.

The existing preamp does have switching noise but it is limited to a very soft high freq click. The switching mechanism in the pre (and probably the main cause of problems) is all electronic using cmos analog switches and ne5534 buffers.

So to cut a long story short, if anyone can give a likely explanation of what is causing the thump, and possible solutions, that would be greatly appreciated

Tony.

[Keith Taylor]

Tony, I am not at all familiar with your amplifier, but make the general observation that most switching thumps are caused by the presence of DC on the sources you are switching to. It could also be caused by an input capacitor having no input terminating resistor to bleed away any charge it may have acquired. This seems unlikely as any half decent amplifier would have taken care of this.
Have a look at the output circuitry of your sources. If there is no coupling capacitor (unlikely) you may have to add one. If there is one make certain that there is a high value bleeder resistor from output to ground.
Keith

[Jan Dupont]

Basically I stuck a DPDT switch in between the 1K output resistors of the preamp and the 1uf input caps on the power amp

It could also be caused by an input capacitor having no input terminating resistor to bleed away any charge it may have acquired.

I think Keith is right on it.
Try put in a 100K resistor from the input side of the power amps cap to ground

[wintermute]

Thanks Guys I'll give it a try, yes certainly with what I have done there is no Bleeder resistor for the direct switched input. There are on the other inputs but this is at the input of the preamp section so I guess that I'm seeing an "accumulation" of dc on the output of the pre while it is switched off. Will be simple to add a couple of resistors, and if it fixes the problem much better than wiring into the speaker relays. Should be able to do this some time today. I'll report on the results

Tony.

[Enzo]

Any time you leave a cap lead floating in the breeze, it will likely generate a pop when it gets connected to something. A high value resistor to ground from that free end should cure that, as suggested above.

[wintermute]

OK well I didn't have any 100K resistors on hand, so used the closest I had which were 68K. Completely fixed the thump for the direct input but made it worse for the preamp input It now thumps both swithching on and off.

I decided to try moving the resistors to the rca jacks on the direct input so at least I'd have one switch movement without problems but this didn't have the desired effect either. happily can switch the direct input on and off without a thump. thumps as expected when going back to preamp, but then thumps again when going to direct...

Decided to check the output of the preamp. It has -1.1V on it!!! No wonder I'm getting a thump!! There used to be a 22uF bipolar where I am taking the feed from, it used to feed into another opamp for the tone controls, followed by yet another for the balance pot... I think I need to investigate why that stage has such a large DC component, or consider putting in a cap between the preamp and the switch...

I'll check the circuit for the preamp and see if there is anything I can do to reduce the DC offset. as that I think would be prefferable to sticking another DC blocking cap in...

Thanks guys

Tony.

[wintermute]

OK well it looks like it isn't as simple as what I might have hoped. The original circuit designers just accepted the offset and stuck in the 22 uF bipolar to get around the problem. I was thinking that maybe I could stick a 47uf cap between the feedback resistor and ground to reduce the offset maybe down to around 170mV (the gain is set at 5.6, so I'm assuming this would work). Still not great but a lot better that 1+ V

It seems the solution is to put a pot in and inject an equal voltage on the input (which doesn't have an offset) which should cancel out the output dc offset. Since my offset is negative I assume I would need to apply this on the positive input.

It's starting to sound like it will be easier to get off my backside and build my new preamp!!

circuit diagram is attached. I'm taking the feed for the amp at the point where the 22uF bipolar electro is. Interestingly I put in a 1K resistor there, I can't remember why, so I have the output of the opamp, 1K resistor, switch, 1uF input cap of the amp. I'm also thinking that 1uF polyester cap is really not going to be blocking the 1.1V -ve dc offset very well!!! The arrow is where I was thinking I could shove in an electro to at least drop the gain to unity at DC.

Tony.

[AndrewT]

Hi,
your output offset = [270k-1k] * input offset current * [4k7/1k0 + 1]
You can reduce the output offset by reducing any or all of the three terms.
Reduce the resistance difference seen by the +in & -in pins. 270k becomes 50k
Reduce the input offset current. FET input opamp.
Reduce the DC gain. Configure as opamp follower or AC coupled.

BTW,
that 0u056F coupling cap kills the bass response. RC~15ms.
Try 0u47F to get RC~120ms. If you go for 50k, the pot should be reduced ~10k and the coupling cap ~2u2F

Why is the comp cap fitted to the 5 & 8 pins?
Was this a follower stage that needed the comp cap for low gain stage?
Who changed the circuit?

[AndrewT]

How about dumping the 5534 buffers and fitting any of the discrete buffers that abound on the Forum?
Start with the VERY SIMPLE, but effective, B1 by PASS.

[Keith Taylor]

Putting a capacitor where you suggest to reduce the DC gain to unity seems a good idea. The reason you would have put a 1K in series with the output is to reduce the phase shift that cable capacitance may introduce at the feedback pick off point were it not there. This can lead to instability. Presumably you have ceramic bypass capacitors near the 5534 supply pins in addition to the electrolytic caps?.

If you have an oscilloscope it may be worth looking at the output, as sometimes what appears to be a DC offset may be very high frequency oscillation with an assymetrically clipped waveform.
Keith