diyAudio (
-   Solid State (
-   -   Buzzing power amp - DC blocker has no effect (

Radioactivebloke 14th September 2012 09:00 PM

Buzzing power amp - DC blocker has no effect
1 Attachment(s)

This is my first post although I've read lots of the threads here.

I use a Peavey PV 3800 power amp to power my big subs. It was silent when I first got it but at some point it started buzzing. The buzzing is constant and appears to be coming from the transformer. I've tried tightening the bolt through it but that didn't help. Eventually I stumbled across several threads here about mains DC offset and DC blocking circuits which reportedly silenced buzzing toroids. I haven't tried to measure the DC offset here but I noticed that the buzzing got louder and quieter when a hairdrier was switched on and off. So I assumed that a DC offset was the problem and built a circuit to block it (see attached image). It is in series with the live wire.

Capacitors: 35V 22000uF (B41231A7229M000 - EPCOS - CAPACITOR, SNAP-IN, 35V, 22000UF | Farnell)
Bridge rectifier: 400V 25A Vf = 0.6 for each diode (CM2504 - MULTICOMP - BRIDGE RECTIFIER, 25A, 400V | Farnell)

If my calculations are correct this should block DC up to about 6A (which is what the label on the back of my amp says).

I finished building it today, plugged in the power amp and... it still buzzes. The buzzing still gets louder when the hairdrier is turned on too. I even tried removing the bridge rectifier so I just had the caps but that didn't help. How can the hairdrier still be affecting the buzzing of my amp? Is there something wrong with my circuit? Or is the hairdrier doing something else to the power supply that could be upsetting my transformer?

Any help would be much appreciated!

Thanks :)

Struth 14th September 2012 10:51 PM

Hi Guys

You definitely want to reconnect the bridge! It is there to protect the caps. As a side note, the bridge itself can be the lowest voltage type offered since it is being wired in a different fashion than its intended use. None of the diodes see more than the diode drop of the antiparallel diodes.

Also, you can reduce the voltage across the caps by Folding the bridge connections one more time so that the ACs are linked and that is one end, and the DCs are linked and that is the other.

Also, you do not have to series-connect the caps. For what it is worth, they can all point the same way but it is customary to have one point one way and the other opposite. This allows lower ESR over all and the use of smaller caps.

The set of caps just requires a ripple current rating equal to or greater than the load current.

However, you have already determined that DC on the mains is not the problem....

The hairdryer noise can get into the system many different ways. Is it to be assumed that only the PV puts out the noise? There could simply be a dirty connection somewhere. If the system has been in place for a while it is good practise to unplug-replug connectors to wipe their surfaces and renew contact.

Have fun
Kevin O'Connor

AndrewT 15th September 2012 08:54 AM

Your diodes will each block ~400mVdc when passing virtually zero current.
Your bridge is set up to give two diode drops in either direction.
The maximum peak current through the caps is limited by a peak voltage of ~800mVpk.

The 4off series parallel caps have an effective capacitance of 22mF.
At 50Hz these caps will have a capacitive reactance of 0.144ohms

With a drive voltage of 800mVpk then maximum (primary) current will be 5.5Apk
That is equivalent to 3.9Aac.

Any transients that exceed the 5.5Apk will "turn on" the diodes. This will happen during start up. It will probably happen if you overload the amplifier.
Hopefully the normal operation of the amplifier will never draw a PRIMARY current of more than 5.5Apk.

Now back to the transformer.
If it buzzes when there is nothing connected to the output, then there is probably some shorted turns inside.
If the transformer primary current (on zero load) is equivalent to more than a very few percent of the maximum rated primary current then your transformer is probably scrap.

Radioactivebloke 16th September 2012 01:02 AM

Thanks for the replies.

Only the transformer in the power amp makes any noise. There is no unwanted hum/noise through the subs it is connected to. I opened it up and unplug-replugged all of the internal plugs I could see but that doesn't seem to have changed anything.

The Vf of the diodes measured at 630mV on my multimeter. I assume that you used 400mV in your calculations because that is approximately the point when the diode will start conducting even a tiny amount? 3.9A should be fine anyway, if it's drawing that much current then I shouldn't be able to hear the buzzing!

Whilst I had the cover off I took a few measurements. The biggest number my multimeter showed when switching on the amp was 0.15A (although I suppose there could be a surge too brief for the meter to register). When measuring zero-load current draw the display alternated between 0.00 and 0.01A. I can't find any detailed specifications for my transformer - the sticker doesn't give a power/current rating - but I'm sure that this is a fraction of the maximum rated primary current.

It does indeed buzz with nothing connected to the output. Is there anything I can do about shorted turns if that is the problem?

I might phone up Noratel (the transformer manufacturer) at some point to ask if they have any suggestions.

AndrewT 16th September 2012 08:41 AM

If the primary has any shorted turns then the transformer is scrap.

If you have the skills and resources required you can save a transformer with shorted turns in the secondary.

To measure the primary current, insert a 10r resistor in the primary feed.
Measure the voltage drop across that resistor. But be warned !!! that resistor is at MAINS VOLTAGE, even if fitted in the Neutral lead.

I use an insulated terminal strip with the resistor wired between two terminals. Then the whole shebang insulated from touching. Just the two screw towers are exposed and the holes are just big enough to insert the voltmeter probes, but no fingers can accidentally get in.

I never recommend any Member use the ammeter function of a DMM.

Struth 17th September 2012 04:24 PM

Hi Guys

A digital meter will be too slow to measure the peak of inrush current into the amp. Assuming this is a conventional split-rail PA, the PT size is basically whatever the power consumption rating is on the back of the amp. Or, if you know that the amp is, say 200Wper ch, then the PT is likely at least 400VA. Anything over 300VA is subject to upset by mains DC. Smaller PTs have enough DC winding resistance to not be troubled by the DC.

The turn-on surge is only limited by the cap ESR and the tiny resistance of wiring, secondary wire resistance, rectifier dynamic impedance - all of which are small. For example, if the amp has +/-50V rails and 1R of total wiring resistance feeding the caps, then the peak turn-on current is 50A. The cap value has no bearing on this, only on how long the surge may last.

Once the caps are charged and the PA is idling - likely at leas than 100mA per channel - then the draw from the mains will be as low as the amp ever draws. This is where DC mains can upset the PT and make it buzz or growl. Under load, the currents are higher and the loaded PT is effectively "immune" to mains DC issues. The DC mains blocker will be doing its job when you are most likely to hear whether it is working or not.

If you modify the DC blocker per my email above, a single cap will perform as well as the four caps you have now. Placing two capos in parallel will reduce the Xc and going to four caps in parallel cuts it even further. As Xc drops, the current that passes through the cap can be increased as long as it is within the cap ripple rating.

Have fun
Kevin O'Connor

kevinkr 17th September 2012 04:33 PM

If there is no shorted turn in this transformer and the low magnetization current of roughly 10mA (0.01A) make this seem likely it is probably a mechanical/QC defect. I've had buzzing transformers when unloaded and this was usually reflective of mediocre quality, loose laminations or windings. (Is this an EI type or toroid? If an EI you might take a look at the lams and how tightly they are clamped) If this was purchased new you might want to return it and ask them for one that is quiet.

You could try diy vacuum impregnation with transformer varnish as a last resort or buy something made by a company that makes quiet transformers for audio use.

Radioactivebloke 19th September 2012 12:29 AM

1 Attachment(s)
Thanks for the input.

The transformer is a toroid. It doesn't have a power rating on the sticker so I can't be sure of the power rating but here's what I know about it:
  • Once I accidentally used the wrong power cord and it blew a 5A fuse so I suppose it's at least 1150VA.
  • The sticker on the back says 6A (6A * 230V = 1380VA)
  • The transformer is 9.7kg which (from comparing to other toroids) I'd estimate to be 1000-1250VA.
Some details from the manual:
  • Rated Power (2 x 4 ohms): 1300 watts per channel @ 1 kHz at <0.05% T.H.D. both channels driven
  • Current Draw @ 1/3 Power (extreme music conditions): 2975 watts @ 4 ohms
  • Current Draw @ 1/8 Power (typical music conditions): 1185 watts @ 4 ohms
Taking all of this into consideration I'd guess that the values of 2975W and 1300W per channel are peak ratings, and that the transformer is probably about 1250VA.

I re-wired the circuit with the capacitors in parallel but there was no noticeable change.

The amp was purchased ex-demo but it's way out of the three-month warranty. Vacuum impregnation looks interesting but I'm worried that it might affect the resale value. The transformer is neatly wrapped up at the moment (see photo) so it would be obvious if I'd tampered with it.

Without knowing the exact specs of the current transformer it will be tricky to replace it. It seems to have an unusual number of windings too - there are a total of 12 wires going into it.

Maybe I should just put up with it until I can afford to buy a better amp!

All times are GMT. The time now is 08:46 PM.

Search Engine Optimisation provided by DragonByte SEO (Pro) - vBulletin Mods & Addons Copyright © 2017 DragonByte Technologies Ltd.
Resources saved on this page: MySQL 18.75%
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2017 DragonByte Technologies Ltd.
Copyright ©1999-2017 diyAudio