Looking to finalize my parts list for a build...basic recreation of Dyna MkIV..and staying with 5AR4 rectifier and capacitor input supply. Knowing the max capacitance of the first cap to be 40uF for this tube, I was looking at 39 uF 630 VDC Solen, or equivalent, but noticed the 330 VAC rating. Pretty sure there's more than that hitting that first cap. Better to stick with an electrolytic here? Thanks in advance.
406DCB450K2J | Illinois Capacitor
https://www.mouser.com/datasheet/2/400/MKP_B32794_798-524911.pdf
https://www.mouser.com/datasheet/2/212/KEM_F3046_C4AE_RADIAL-1107630.pdf
??
High current pulse capability polypropylene film capacitors (for motor run use, presumably).
- Very low ESR (0.0071 ohms for the Illinois cap, 0.0028 ohms at 10kHz for the KEMET cap) means it wouldn't heat up so much from large current AC pulses, correct?
- Ripple Current: 11 A at 10 kHz and 70°C (for the Illinois cap) - I'm not sure how that equates to ripple current at 120Hz (or 100Hz).
- These 40uF 500V film caps cost around $15 each.
Looking at the equivalent electrolytic, taking a KEMET 47uF 450V rated part designed for high ripple current:
- Its ripple current rating is 0.34A at 120Hz.
- Its ESR looks like it's about 1.8 ohms.
https://www.mouser.com/datasheet/2/212/KEM_A4008_ESG-1101971.pdf
- This 40uF 450V electrolytic capacitor costs $2.35. (That's a big difference!)
I wouldn't say electrolytics are not up to the job. They most certainly are. They're designed for that job. But I'm not sure typical 450V-rated electrolytic power supply caps have actually "better" ripple current ratings than motor run (film) caps, which are also designed to be used under high ripple current and AC pulse conditions.
For the reservoir capacitor, if instead of a good electrolytic capacitor one used a plastic film capacitor designed for high ripple current conditions, I don't know if that would improve the power supply performance enough to be audible. Technically speaking, it should improve things. But as has been pointed out, it would make more a difference (possibly even audible) if you put the motor run (film) cap as the last decoupling cap, right before the load.
What if you took an expensive audiophool MusiCap or whatever and used it as the reservoir cap? I have no idea. In a 39uF 450V rated part, something like a MusiCap would most certainly be a gigantic waste of money.
--
https://www.mouser.com/datasheet/2/400/MKP_B32794_798-524911.pdf
https://www.mouser.com/datasheet/2/212/KEM_F3046_C4AE_RADIAL-1107630.pdf
??
High current pulse capability polypropylene film capacitors (for motor run use, presumably).
- Very low ESR (0.0071 ohms for the Illinois cap, 0.0028 ohms at 10kHz for the KEMET cap) means it wouldn't heat up so much from large current AC pulses, correct?
- Ripple Current: 11 A at 10 kHz and 70°C (for the Illinois cap) - I'm not sure how that equates to ripple current at 120Hz (or 100Hz).
- These 40uF 500V film caps cost around $15 each.
Looking at the equivalent electrolytic, taking a KEMET 47uF 450V rated part designed for high ripple current:
- Its ripple current rating is 0.34A at 120Hz.
- Its ESR looks like it's about 1.8 ohms.
https://www.mouser.com/datasheet/2/212/KEM_A4008_ESG-1101971.pdf
- This 40uF 450V electrolytic capacitor costs $2.35. (That's a big difference!)
I wouldn't say electrolytics are not up to the job. They most certainly are. They're designed for that job. But I'm not sure typical 450V-rated electrolytic power supply caps have actually "better" ripple current ratings than motor run (film) caps, which are also designed to be used under high ripple current and AC pulse conditions.
For the reservoir capacitor, if instead of a good electrolytic capacitor one used a plastic film capacitor designed for high ripple current conditions, I don't know if that would improve the power supply performance enough to be audible. Technically speaking, it should improve things. But as has been pointed out, it would make more a difference (possibly even audible) if you put the motor run (film) cap as the last decoupling cap, right before the load.
What if you took an expensive audiophool MusiCap or whatever and used it as the reservoir cap? I have no idea. In a 39uF 450V rated part, something like a MusiCap would most certainly be a gigantic waste of money.
--
Last edited:
Initially the plan was 2 Sprague TVA Atom 20uF/600v in parallel for the input caps, which is what I used in a pair of MkIII's years ago. These aren't exactly cheap, even at Mouser. Then I took a look at film caps...KEMET polypropylene 40uF/600v that cost less and took up about half the space. I looked at these solely for space saving...not for audio quality. I started laying out a chassis plan using a graphics program to avoid surprises and find the space to place wiring carefully. The "bowl of spaghetti" or rats nest wiring approach hurts my eyes to look at. I can make the 2 TVA Atoms fit, but the KEMET's would give me a little extra space. I just got put off by the VAC rating.
Last edited:
I bought a box of 45uF polypropylene motor run caps good for about 600vdc and they were only about $4 each. Great cap for the money and should out last any electrolytic. I have only used two since I got them so I got a like 10 left. The two I used were decouple caps at the node feeding my output transformers in my single ended amp, I think it improved the sound over the electrolytic I tried. I have been toying around with using them exclusively in the power supply of my latest single ended amp, parallel 4P1L wired as triodes. Gonna need a large chassis or separate PSU chassis.
DF96 is correct, there won't be much AC voltage on the power supply cap, the reservoir cap will have the largest 120Hz AC pulse but it will be nowhere near 330v.
DF96 is correct, there won't be much AC voltage on the power supply cap, the reservoir cap will have the largest 120Hz AC pulse but it will be nowhere near 330v.
They may seem rather large, but I find ASC X386s 440VAC Motor Run Film capacitors to be excellent. They do come in 40uF as well.
If space is a concern, Vishay MKP 1848 DC Link Film capacitors might be an option. I purchased some 40uF that are rated up to 900VDC and cost less than 4 EUR. The sound is a bit more forward than the ASC Motor Run's but also excellent. Don't feel limited to the Vishay's though - there are a TON of DC Link capacitors out there to try. I just limit myself to ones that are not too pricy...
Now, you might still want to go for electrolytics because you have them on hand, etc. but don't forget the possibility of by-passing the electrolytic with a film cap. For this, I have been using Russian PTFE K72P capacitors. People seem to 'notice' the difference.
If space is a concern, Vishay MKP 1848 DC Link Film capacitors might be an option. I purchased some 40uF that are rated up to 900VDC and cost less than 4 EUR. The sound is a bit more forward than the ASC Motor Run's but also excellent. Don't feel limited to the Vishay's though - there are a TON of DC Link capacitors out there to try. I just limit myself to ones that are not too pricy...
Now, you might still want to go for electrolytics because you have them on hand, etc. but don't forget the possibility of by-passing the electrolytic with a film cap. For this, I have been using Russian PTFE K72P capacitors. People seem to 'notice' the difference.
Last edited:
I use PS film caps all the time.
The Kemet brand you referred to are pretty good on my SE amp. If you check around you can find that they have various voltages up to and over 1000vdc. I used mostly Solen before that but have been using a lot of the DC link box caps too, mostly Panasonic. You need to solder wire leads and i like to use 18 awg solid core for easy layout.
The Kemet brand you referred to are pretty good on my SE amp. If you check around you can find that they have various voltages up to and over 1000vdc. I used mostly Solen before that but have been using a lot of the DC link box caps too, mostly Panasonic. You need to solder wire leads and i like to use 18 awg solid core for easy layout.
I would not use a 450V rated cap on a Dyna Mark IV B+ supply.
The manual states that the GZ-34 / 5AR4 plates are 370VAC (523V peak). The GZ-34 / 5AR4 output of a working unit is 440VDC (that is with both EL34 tubes drawing current, and includes the GZ-34 / 5AR4 voltage drop With That Current Draw). If one of the EL34 tubes dies, the B+ will rise to Way over 450VDC, Ouch!
The manual states that the GZ-34 / 5AR4 plates are 370VAC (523V peak). The GZ-34 / 5AR4 output of a working unit is 440VDC (that is with both EL34 tubes drawing current, and includes the GZ-34 / 5AR4 voltage drop With That Current Draw). If one of the EL34 tubes dies, the B+ will rise to Way over 450VDC, Ouch!
Why someone would worry about the finite life of electrolytics I can understand to a point (I have 50+ year old electrolytics that work just fine).
In the context of PSU impedance, why anyone would worry about the ESR of an electrolytic cap, and proceed with a tube rectifier, is far beyond me. Am I missing something?
In the context of PSU impedance, why anyone would worry about the ESR of an electrolytic cap, and proceed with a tube rectifier, is far beyond me. Am I missing something?
Nope...that isn't gonna happen. First cap after the rectifier would be rated at 600v.
Who me? I'm not worried about anything now. I'll put that KEMET cap right where I wanted, save a little space, everything will be fine. And as far as the 5AR4 goes, as long as you don't ask too much of it, it will be fine. I don't need a ton of power...being an apartment dweller. 30 wpc or so is more than I need..or can use. I'm just feeling the need to do a resto-mod with the MkIV circuit. I had a pair about 20 years ago and I enjoyed them a lot.
Why such large capacities behind a tube rectifier? If there is an LC stage behind it, there is no need to strain the rectifier this way. Output tubes are not as sensitive to hum on the B+ and the gain stages are generally behind an RC network in the B+ to lower ripple considerably.
e-caps suck in general, try to avoid them when possible. MKP or oil caps at these values are not expensive, and you can often find a lot of them second hand. ESR is hardly an issue here as we don't have 4-5A ripple current like on low-voltage high-current supplies. Also the difference between a "regular" and "low esr" cap is not that great that it changes anything in the load of a tube rectifier.
e-caps suck in general, try to avoid them when possible. MKP or oil caps at these values are not expensive, and you can often find a lot of them second hand. ESR is hardly an issue here as we don't have 4-5A ripple current like on low-voltage high-current supplies. Also the difference between a "regular" and "low esr" cap is not that great that it changes anything in the load of a tube rectifier.
Hmmm...just downloaded that datasheet. I've never seen any resistors used in that position in stock Dyna gear. Wonder if there's enough DCR already in the stock power trans. The "plan" was to use a pair of EDCOR XPWR-002 and the DCR spec is not listed. Guess I'll have to wait and measure when they arrive. Thanks for the heads up.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.