Zen Enlightenment.

I have not yet reached total enlightenment with Zen.

I acquired all the new parts, and it's together once again.
This time with 100V MOSFETS, a 1.5mH coil and a 5W R1.

I powered it up and measured the positive voltage at 35.8VDC in the circuit.

I tried to adjust the pot to get 17.9VDC at the drains as the Zen documentation indicates.

My problem is, I turned the pot all the way one way and only was able to get the voltage to 20.5V.

It looks like R8 and the pot act as a voltage divider circuit feeding the Gate. So, using math, a 25Kohm resistor would only bring the voltage down to 23.45V. Close.
I need more like a 50Kohm pot in there to adjust it right.

Anyone disagree with this math? Is there another explanation?

I haven't done any other testing. This time it became warm after only a minute or so of idling. So, I wasn't going to mess with it.

Also, I don't have a frequency generator. Anyone know of a good DIY generator? For audio, it wouldn't have to be great, just 20Hz to 20KHz sine waves.

Thanks,
 

vdi_nenna

Member
Paid Member
2000-10-10 7:27 pm
PA, USA
re: Zen Enlightenment

Pixie, Sounds like you're almost there!

Did you check for DC at the output? If you have little (less than a volt)or no DC at the channel outs, try connecting an old 8 ohm speaker. Run some low level music or test tone off a cd through it.

For what I remember, the voltage measured should be half the supply voltage.

The value for the pot was different in the 2 versions...did you use the right one...just to be sure.

It gets warm right away. That sounds normal to me.

Because of the size of the caps, I don't even think the Zen amp does 20hz to 20 khz.
 
Yes, I'm using the larger pot. The original was 10K the new one is 25K. But, my theory is that is still not big enough. I think it should be more like 40-50K.
I don't want to cut traces to add a resistor. So, I guess I'll trek over to Radio Shack to find a larger pot.

My output is connected to one of those 8ohm 100W resistors used for bench testing amplifiers. I picked it up from PartsExpress.

I don't want to connect a signal and try to drive this thing until I get the MOSFETs balanced.

I don't have a good test CD. The one I have only has sub-woofer test tones and they vary wildly in amplitude. So, I can't test the positive and negative clipping symmetry. As was suggested in my last thread.

You're right, the output shown in the document only goes up to 15KHz. That's OK with me. I'm going to drive a center channel with it.
 

vdi_nenna

Member
Paid Member
2000-10-10 7:27 pm
PA, USA
A dummy load is a good thing to have.

I rebuilt a set of speaker crossovers for a frind last night. I did some tone testing at -10 db reference. After 15khz, I couldn't hear a thing. 18khz wasn't even there.

Anyway, I think you're almost there. If it didn't catch fire this time and there is no DC at the output there isn't too much more to it. It's just too damn simple.

good luck man.
 

FYC

Member
2001-01-04 10:29 am
Hong Kong
Something fishy

I do not believe that going higher in VR value will help.

Look into the schematics, the 25K VR and the 47k resistor form a voltage divider to get the grid bias voltage, which should be 4V. Assuming you need 18V at the drain, the ratio is 4/18=0.22, so the VR should be tuned to about 13k. Even if grid is 5V, 5/18=0.27 and VR should be about 17k. The 25K VR will work out.

I would suggest you recheck your wiring, worst case is your MOSFET is gone!?
 
OK. I replaced the pot with a 100K. It worked. I was able to adjust the drain voltage to 18V. After a little burn-in the +V settled at 36V exactly. It ran great for about 20 minutes. It had no DC at the output and 0.000VAC at the output. Great, right?

Well, after it didn't blow up, I put a 200Hz sine through it. It worked, kind of. I borrowed an ocilloscope from a friend. The output signal looked very dirty. Then about every 5-6 seconds it would ouput just garbage. I realized I was measuring about 20mV signal. So, I turned the output up and my 10A fuse at the +V blew. It wasn't to clipping yet, and the signal quality hadn't improved.

I'll have to recalculate the amperes required from the power supply. I thought it wouldn't exceed 10A.

That aside, what was all that noise? I wouldn't dare connect this thing to a speaker.

OK. I'm getting frustrated now. What could I be doing that's so wrong?

Thanks,
 
How about the zener? Does anyone else think it may be breaking down? That would explain a time delay between the bursts of noise. A time delay would also be explained by capacitors.

When it does this there is no recognizeble signal at the output. It's total garbage.

I think it's just time to replace all the parts and start over.

Any other suggestions before I scrap it?

Thanks,
 

vdi_nenna

Member
Paid Member
2000-10-10 7:27 pm
PA, USA
Take all the parts you have and test them out of circuit.

You can test the FETS w/ a little circuit from the Pass A75 amp project. All your really need is a 15v supply and hardwire a small perf board from the schematic shown.

Get a 25+25v @ 6 amp trans. or figure out how to reduce the the secondaries after the rectifiers to 34v+34v from your 30+30v transformer. Then get the Zen board from AudioXpress for 12 to 15 bucks.



There are some things you need to know about putting the transistors on the board, but I can help you with that if you want. I've done it wrong enough times and finally done it right once, and it's worth it.
Don't give up, re-group and start over when you can.

That's my advise. Take it for what it's worth.
 

FYC

Member
2001-01-04 10:29 am
Hong Kong
There must be something wrong if you need to use a 100k VR to get to 18V at the drain. With what you describe, certainly you need to recheck your complete wiring and your MOSFET. MOSFET are easily killed by statics, or it run into oscillation due to connection problems.

Don't give up, my Zen amp work perfectly immediately after hook up, and was so far so good. It is not a terribly difficult project.
 
Please, stop saying it's not a difficult project!

This is my first analog circuit!
To prove my predisposition to digital, I just remembered yesterday that the G is for Grid not Gate. Sad.

Thanks for the help though. It's appreciated. I wouldn't have gotten this far without you guys. Especially Vince. Thanks man.

I think I will get that circuit board. I'm going to keep the power supply part separate though. My father made me an aluminum case that has a separate section for the power supply. And, I'll remember to check the diagram for the connection of the small transistor. I'm going to start over with new FETs. So, I'll be starting with a clean slate.

Thanks,
pixie
 
Hi Pixie et al - hope you got it going! Others said it's not difficult - I agree - your probs have got to be due to zapped mosfets, bad wiring, etc. Give it another go - hand check all the bits like vdi_nenna said that could cook before inserting, wire mosfet leads & remove only when wired the rest (is this overkill?). Resistors - cute new(?) coding scheme - are you getting them right - DON'T always rely on the supplier doing so either).

I'm building beefed up Zen Revisited - 4amp bias, 50v psu, twin paralleled channels working into 4ohm spkrs. 2 of 4 completed channels worked first time - but you need to set bias point (as advised by NP) for symetrical clipping ('scope) - it's NOT 1/2 supply volts - rather more like 60% - that cooks the lower mosfet more too - beware the h/sinks.

I'll report results (& any more hints) in due course... Not put these 2 together yet... Most difficult (in a domestic environment) is the heatsink with these things. 80% of my effort - high bias, high volts - VERY big h/s :)( Great results :))

Pre-amp will be BOZ - NP says it'll be fine into the 500ohm i/p (2 parallel amps). BOZ as published (5K o/p pot) will have a very skewed volume curve into Zen Revisited with it's much lower i/p impedance too. I've worked out stepped attn. resistor vals to hopefully match the two - we'll see..

Cheers to all

Pete