DIY Audio Projects Forum

300B SET Design Project
Page 49 of 53

Author:  ILoveHiFi [ 30 Jun 2019, 22:56 ]
Post subject:  Re: 300B SET Design Project

I looked at your PDF it says
Series to shunt reactance ratio check
implications of not being >= 20?

The only thing that matters is that the circuit has enough capacitance to ground to bypass the serires resistance.
If the cap + series resistance yeilds a cut off frequency of much lower than 20 hz then your fine

Author:  Woodo [ 02 Jul 2019, 22:37 ]
Post subject:  Re: 300B SET Design Project

ILoveHiFi wrote:
I looked at your PDF it says
Series to shunt reactance ratio check
implications of not being >= 20?

The only thing that matters is that the circuit has enough capacitance to ground to bypass the serires resistance.
If the cap + series resistance yeilds a cut off frequency of much lower than 20 hz then your fine

Ok, I haven't quite got it in my head yet how it's supposed to work quantitatively but I'm guessing the resistance/reactance plays a part in selectively shunting the 100 Hz ripple to ground while allowing the "pure" dc to pass. If the ratio is 1 or near one does that then mean half the ripple will not be bypassed?
My JJ5ar4, 22uf cap and 100R and 10kR to drop voltage to the 6sn7 arrived. I thought I might try the 5ar4 in place of the 5u4g first and check hum. If necessary to quell any inrush current I could try using the 22uf cap in place of the 39uf cap. The design guide calculates a higher amount of ripple then to deal with. So then if necessary I could try the 100R and 39Cap which cross fingers results in the target ripple of 12 mV. The data sheet for the 5ar4 refers to having voltage dependent series resistance in the transformer secondaries . I need to check the amount but either 125R or 175R per plate. I'm unsure whether I would need this for my domestic amplifier but if so then of course I will add it. In the power supply design guide I interpreted that this would be required only if hot switching is used. Not sure what this exactly means but I don't intend to use it intentionally.


Author:  ILoveHiFi [ 02 Jul 2019, 23:13 ]
Post subject:  Re: 300B SET Design Project

Inrush current dosen't matter, the transformer or valve rectifiers will only be damaged when they are over power ratings for continous time.
Few minutes over power wouldn't damage them.

You don't have a inductive load so all you need to care is the series resistance and capacitance to ground forms a RC filter with cut off frequency much lower than 20HZ
The only inductive filtering is the series transformer filter and thats the first stage, this will not effect the following stages of RC filter.
The load only becomes inductive when you have audio swining (output transformer), you only get the noise when its on idle. Once you start playing music then the noise is going to be negeltible anyways.

And you do not need to check if the actuall filter has the poles in RHP because its virtually impossible to have RHP poles with the thing your currently doing.
(Its impossible to have unstable filtering system that will oscillate over time)

Author:  Woodo [ 04 Jul 2019, 16:16 ]
Post subject:  Re: 300B SET Design Project

I swapped the 5u4g for the 5ar4 as per Matt’s circuit. I haven’t made any other changes yet such as lowering the first capacitor value or put resistors in in the secondary wires. B+ increased to 447v and 270v for the 6sn7’s. Although same hum/buzz remains. I put my ear to the outside of the power xfmr and I’m fairly certain there’s a similar noise coming from there.

Author:  Woodo [ 06 Jul 2019, 20:43 ]
Post subject:  Re: 300B SET Design Project

I swapped the 39 uf first capacitor with the 22 uf capacitor. I didn't notice any change in the hum/buzz level but have left this cap in for now anyway. I added 100 ohm resistors to the secondary leads to the rectifier. This would bring secondary resistance up to 74+100=174 ohms which is likely more than necessary, but no expert in this matter for sure. if I read the JJ data sheet correctly ( ... /gz34s.pdf) 125 ohms per plate or so is required but not certain if this is total secondary resistance or additional. However, there was no change to the buzz from adding the resistance so I must have a different problem there. B+ has dropped to 396v and back to about 240v for the 6sn7's. The transformer appears to remain cooler after about 1 hour of music playing so this might be doing some good. Sound was still great to me, perhaps subjectively less attacking and dramatic on some passages, hard to judge.

One thing I noticed with one of the tracks I played, "welcome to the machine" on "wish you're here" by pink floyd the opening section has some low frequency synthesised bass which I haven't heard on my little 6.5 inch two-way speakers since I had the subwoofer connected to my class d pioneer theatre amp. The speaker specs show roll-off around 45 hz. With this version of the valve/vacuum tube amp and no subwoofer the little woofers were putting out some reasonable output on that passage, enough not to worry about having a subwoofer.

Author:  Woodo [ 12 Jul 2019, 08:09 ]
Post subject:  Re: 300B SET Design Project

Song excerpt. Sorry couldn’t post longer sample. file size restriction.

Author:  ILoveHiFi [ 12 Jul 2019, 16:16 ]
Post subject:  Re: 300B SET Design Project

Matt hasn't posted for a long time and not sure why you want to thank him.
If you dislike me and thats the case don't worry.

If you don't like taking advice and you still couldn't fix the noise then thats your prolbem. I've littratley told you every area that is possible to get it wrong.

Author:  Woodo [ 12 Jul 2019, 21:18 ]
Post subject:  Re: 300B SET Design Project

I appreciate all and any advice I receive on DIY Koichi-San particularly yours so sumimasen if I haven't explicitly thanked you so a big shout of THANKS to you too and I apologise for all my noob questions. I really enjoy reading your posts, you are obviously passionate about hifi and contribute a great deal. I have a knowledge and experience gap to implement the advice of people who obviously know a lot more but I think I have tried most things you suggested. The hum issue I haven't resolved completely yet, probably a bad solder joint or my compromised layout (I won't make this mistake on the next build) or something else I have overlooked. I am reading through all the 100's of posts about hum to see if there is something that resembles my issue. I'll get there in the end somehow. I was thanking Matt really for this awesome design because hum aside, which is only noticeable in quiet passages, I can't get over how detailed and articulate the amp sounds on some of the music I have.

Author:  ILoveHiFi [ 12 Jul 2019, 23:26 ]
Post subject:  Re: 300B SET Design Project

Hey sorry, I've been under allot of pressure latley with a lot of work load.

But with what you've tired, increacing capacitance and reduce the power supply noise.
I don't think it made any diffrence.
But if the noise reduced quite noticibliy when you added the capacitors, try adding more capacitors.
If the noise is still auditable during low passage something is quite wrong.

The original design may be wrong but looking at it I don't see any prolbems

So the conclusion I'm at is that your layout is wrong or the ground wiering is wrong.
I belive the noise can be solved if you fix the layout or ground wiering.
I didn't say anything untill now.
Since you didn't ask any further previously and I thought it was already obvious that the ground or layout is wrong since had mentioned previosly.

Or that the heater could be wrong
The two heater 100R resistors are connected to the ground of the power amp, which is also connected to input ground.
If you get this wrong you will get some heater noise

For the layout is the voltage transformer and the filtering transformer for power supply needs to be placed in a way that is away from all signal path.
They radiate magnetic fields and this will directly flow into the amplifier.

The exposed coils radiate field and creates noise. Metal wall connected to ground of amp can sheild the noise.
Or you can rotate the transformer 90 degrees and it radiates in a diffrent direction.
The part where is the metal core of the transfomer that shouldn't radiate much noise.

The ground is bit harder to explain, but for a start is that if you connect the ground to flitering capaictor ground then you will get hum.
Ground to mains needs to be somewhere at or near the ground return point which is local and close to the amplifier section
And you only have one wire going from the B+ ground to amplifier ground to get the voltage.

The wires from transformer should be placed near the filtering capacitors. Then only two wires to get the voltage and ground goes to amp.
The way to fix layout ground is to trial and error. Don't be affriad to make changes gamble a bit, you can always revert to the old one.

Author:  ILoveHiFi [ 12 Jul 2019, 23:45 ]
Post subject:  Re: 300B SET Design Project

Also be aware of the ground connection from pre amp, if anyone is wrong then the power amp will not hum when nothign is connected. But once you connect the pre amp you start humming.

Power by batteries or having voltage regulator is a easy way to cover up the ground wiering prolbem. Which I saw was the way some people on this from solved their probelms.
However the ground wiering prolbem will still reduce sound quality even if it dosen't hum.

Also you cannot connect amp ground to chassis, amp ground has to be directly to 230v ground pin on mains with a fairly thick gauge wire.
Then on the 230v ground pin from walls you connect a seperate wire to chassis directly.
Some countries have strict regulations and ground lifting or grounding via resistor diode or capacitors are illegal. It must be connected with a low resistance wire to mains ground.

Page 49 of 53 All times are UTC - 6 hours [ DST ]
Powered by phpBB® Forum Software © phpBB Group