I edited my simulation with the new values of DCR, and I get exactly the same values as those you measured.
Not much. I believe that with the new resistor values it works well enough.
Warm Audio WA-67 - Teardown
- Thread starter Whoops
- Start date
Help Support GroupDIY Audio Forum:
So what can be the consequences of having a clipped AC waveform on B+?
Wouldn't it affect performance?
Sorry if these are basic questions to you, this is new stuff to me so I'm just learning.
I'm also using this teardown to explore and learn more about some aspects I never investigated before
Thanks for your patience
How do you define "performance"?
Probably as you or most of us would define it
To me it's an entirely abstract term (in and of itself). You first used it, so i asked what YOU meant by it. Hence the "how do YOU" in the question there 
Since @abbey road d enfer already drew up the simulation, we should ask about the results of an AC sim, and see how low the cutoff frequency of that multistage RC lowpass filter is. With those humongous values, i'm reasonably sure you could even have a full squarewave at the input, and still have quiet DC at the output...
Since @abbey road d enfer already drew up the simulation, we should ask about the results of an AC sim, and see how low the cutoff frequency of that multistage RC lowpass filter is. With those humongous values, i'm reasonably sure you could even have a full squarewave at the input, and still have quiet DC at the output...
Actually, the waveform is always clipped somewhat, maybe here a little more than usual.
As long as the DC voltages are correct and the residual ripple adequate, there is nothing wrong with it.
![Electronics Soldering Iron Kit, [Upgraded] Soldering Iron 110V 90W LCD Digital Portable Soldering Kit 180-480℃(356-896℉), Welding Tool with ON/OFF Switch, Auto-sleep, Thermostatic Design](https://m.media-amazon.com/images/I/41gRDnlyfJS._SL500_.jpg)
No it’s not, it’s helping me because I asked and helping any other WA67 owners and help other people interested in this mic or U67 circuits that might read it in the Future.
It is no different than any other thread in this forum.
You already said in this thread that you were not interested in helping, that’s fine, just don’t post please
You are not obliged to post just move on
Thank you
Last edited:
Thank you so much for the explanation Abbey.
So I will do as you suggested (moving and adding Zeners) and consider the PSU done
I will also change the 4,7uf C9 for an higher voltage capacitor, and will try also other values here, from 4,7uf to the u67 stock values of 1uf and 0.5uf, and decide which value I think sounds the best.
Will try also a K67 capsule instead of the “flatter” stock capsule, but I would like to do that when I have a Neumann U67 side by side with the WA67 so I could compare both and decide. I have to borrow a U67 and do that
T can't really comment on that, since the value of the cap reacts with the xfmr, which is probably different than the Neumann, and sonic preferences are undebatable.
Yes for sure,
Will try it out and check the difference it makes or not and which one I prefer.
Why not try the higher voltage zener chain right after C2 instead of way down the line towards the load ?
You can always bleed off a little more current with a resistor down the line to achieve a particular voltage at the anode if thats what you want ,it could also serve as a path for the caps to discharge through after powerdown . I see no reason not to increase the value of caps C2 on to maybe 220 or 470 uf , but no need to increase C1 as it could make the peak ac current drawn from the transformer worse , if you want to try and reduce peak current and the chance of saturation you might even reduce C1 to good effect ,although resulting DC voltage will drop a bit ,so the zener/resistors might need to be adjusted .
I like the fact that it takes minutes for large caps (470uf)to charge through kohm resistors , with a zener early on in the chain that ramp up is likely to be even slower , it takes a good 20-30 minutes for a tube mic to stabilise in temperature terms (components and casework) , I would'nt use it to record sound until its reached that point in any case , I dont mind if the HT builds over the same time frame as it takes for the mic to reach operating temperature . PSUD II allows nice simuluations of tube type ht supplies and you can show expected ramp up times . Ive found the results to be in agreement with subsequent real circuit builds .
I have one power supply I use for preamps , it takes at least 3 minutes for the HT voltage to ramp up on load , it includes an EZ81 , signal tube heaters have have a good 30 second lead time before the ht even gets above 20 volts dc and it just rises gently until it reaches its maximum a few minutes later . That supplies 30ma or so HT current.
The PSUD II software can be found on way back machine , under duncans amps , I downloaded a version dating back to 2018 .
You can always bleed off a little more current with a resistor down the line to achieve a particular voltage at the anode if thats what you want ,it could also serve as a path for the caps to discharge through after powerdown . I see no reason not to increase the value of caps C2 on to maybe 220 or 470 uf , but no need to increase C1 as it could make the peak ac current drawn from the transformer worse , if you want to try and reduce peak current and the chance of saturation you might even reduce C1 to good effect ,although resulting DC voltage will drop a bit ,so the zener/resistors might need to be adjusted .
I like the fact that it takes minutes for large caps (470uf)to charge through kohm resistors , with a zener early on in the chain that ramp up is likely to be even slower , it takes a good 20-30 minutes for a tube mic to stabilise in temperature terms (components and casework) , I would'nt use it to record sound until its reached that point in any case , I dont mind if the HT builds over the same time frame as it takes for the mic to reach operating temperature . PSUD II allows nice simuluations of tube type ht supplies and you can show expected ramp up times . Ive found the results to be in agreement with subsequent real circuit builds .
I have one power supply I use for preamps , it takes at least 3 minutes for the HT voltage to ramp up on load , it includes an EZ81 , signal tube heaters have have a good 30 second lead time before the ht even gets above 20 volts dc and it just rises gently until it reaches its maximum a few minutes later . That supplies 30ma or so HT current.
The PSUD II software can be found on way back machine , under duncans amps , I downloaded a version dating back to 2018 .
Last edited:
soapfoot
Well-known member
There would be no significant advantahe in terms of noise, but the dissipation in the resistors that preced the zeners may increase, particularly in the case of a hot mains voltage. OTOH, the zeners may have a better regulation with increased current, not that it really matters.
I ditched PSUD 2 a long time ago. It doesn't offer the same flexibility as LTspice, in particular I don't see any possibility to insert a Zener, and I find the UI somewhat deterring.
So if you increase your last cap(C6) to 680uf , its around 110 seconds before you hit the target of 212 volts
Its possible in PSUD to add in the effect of the zener diode regulator in the form of a current sink , If I knew what the zeners were pulling current wise its easy to do . I'll give it one more try after a coffee .
In the two last examples with two 680 uf caps in positions C5 and C6 I estimate its around 155 seconds before you hit 210 volts DC on the output , as you can see the voltage at C6 comes up lovely and gently , theres nothing for about two seconds , after 10 seconds its still less than 40 volts across the HT line.
Edit PSU startup graphic replaced showing volts/time for C1-C6
Its possible in PSUD to add in the effect of the zener diode regulator in the form of a current sink , If I knew what the zeners were pulling current wise its easy to do . I'll give it one more try after a coffee .
In the two last examples with two 680 uf caps in positions C5 and C6 I estimate its around 155 seconds before you hit 210 volts DC on the output , as you can see the voltage at C6 comes up lovely and gently , theres nothing for about two seconds , after 10 seconds its still less than 40 volts across the HT line.
Edit PSU startup graphic replaced showing volts/time for C1-C6
Attachments
Last edited:
Note I did also change resistor R1 value to 2.7k in the sim to get a better match to your voltages , I chose to dump 3ma in the current tap (Zener) because its approximately what the Neumann circut does with the 68k resistor .
Looks like you want a zener chain with 244 volts and at least 1.3W dissipation , I did the calculation here ,
calls for 2750 ohm resistor ,so things line up well with PSUD.
https://www.petervis.com/electronics guides/calculators/zener/zener.html
Looks like you want a zener chain with 244 volts and at least 1.3W dissipation , I did the calculation here ,
calls for 2750 ohm resistor ,so things line up well with PSUD.
https://www.petervis.com/electronics guides/calculators/zener/zener.html
Attachments
Last edited:
The problems are easy to fix if you know anything about electronics.
Really whoops you are giving me crap after I posted help in this thread
I think people should look at the AT and Rode(there is a good one or maybe more) and MXL(there is a good one that I have three of) and some other microphones that people miss.
Then there are the 251s like "clones" with stupid relays for switching.
Really whoops you are giving me crap after I posted help in this thread
I think people should look at the AT and Rode(there is a good one or maybe more) and MXL(there is a good one that I have three of) and some other microphones that people miss.
Then there are the 251s like "clones" with stupid relays for switching.
And that's the big problem here. No educated guess can solve an exponential equation.
LTspice simulations predict much less than 1mA in the zeners. Considering the mic draws less than 1mA, taht's a significant error.
How much is the current in the microphone going to vary depending on signal level ?
maybe 3% of 1ma with class A ? Our zener doesnt have much work to do to compensate for current in the circuit itself ,
What I want to know is how to set the bias point for the zener diodes so we make best use of their abillity to regulate any small mains voltage variations ?
Would it be more effecient to arrange so the zener barely conducts under normal circuit opperation ,maybe 0.1mA , and only in the case of mains over voltage does or a disconnected load does it start to dump more current ?
I tried a few simulations of the NU67 psu with 3x 680uf caps and 68k bleed resistor at the end of the chain , it gives a nice smooth ramp up .
On the question of where to put the zener in the circuit though , towards the source or adjacent to the load , Im no wiser .
maybe 3% of 1ma with class A ? Our zener doesnt have much work to do to compensate for current in the circuit itself ,
What I want to know is how to set the bias point for the zener diodes so we make best use of their abillity to regulate any small mains voltage variations ?
Would it be more effecient to arrange so the zener barely conducts under normal circuit opperation ,maybe 0.1mA , and only in the case of mains over voltage does or a disconnected load does it start to dump more current ?
I tried a few simulations of the NU67 psu with 3x 680uf caps and 68k bleed resistor at the end of the chain , it gives a nice smooth ramp up .
On the question of where to put the zener in the circuit though , towards the source or adjacent to the load , Im no wiser .
Similar threads
