People, let's concentrate on the facts.
1. Here, the rectifier tube does not rectify anything. It's just a big, expensive NTC thermistor. Its purpose is to delay high voltage in order to protect the other tubes.
So, this rectifier MUST be indirectly heated, or else it starts conducting too early, defeating its purpose.
2. There is no reason to change this tube for a NOS one: it does not rectify - there will be NO benefit in sound quality!
3. Chinese 5Y3 is indirectly heated, while western 5Y3 are directly heated! They are not the same tubes!!
4. In any case, for a rectifier tube you must respect the total minimum series resistance, before and after the tube, and you should respect the total recommended maximum capacitance too. 270µ is quite a lot...
The 5Y3 accepts 2.5A for 0.2 seconds; here, the peak current is closer to 3A.
5. The Chinese probably wired the power transformer for 220V. Here, the mains are around 235V. It means that every voltage in this amp is much higher than normal, including the heaters!!! The tubes will not appreciate it, and the rectifier will flash more easily.
1. Here, the rectifier tube does not rectify anything. It's just a big, expensive NTC thermistor. Its purpose is to delay high voltage in order to protect the other tubes.
So, this rectifier MUST be indirectly heated, or else it starts conducting too early, defeating its purpose.
2. There is no reason to change this tube for a NOS one: it does not rectify - there will be NO benefit in sound quality!
3. Chinese 5Y3 is indirectly heated, while western 5Y3 are directly heated! They are not the same tubes!!
4. In any case, for a rectifier tube you must respect the total minimum series resistance, before and after the tube, and you should respect the total recommended maximum capacitance too. 270µ is quite a lot...
The 5Y3 accepts 2.5A for 0.2 seconds; here, the peak current is closer to 3A.
5. The Chinese probably wired the power transformer for 220V. Here, the mains are around 235V. It means that every voltage in this amp is much higher than normal, including the heaters!!! The tubes will not appreciate it, and the rectifier will flash more easily.
"1. Here, the rectifier tube does not rectify anything. It's just a big, expensive NTC thermistor. Its purpose is to delay high voltage in order to protect the other tubes.
So, this rectifier MUST be indirectly heated, or else it starts conducting too early, defeating its purpose."
Protect? Why does every commercial and very successful manufacturer not use anything like you suggest in your schematic? Have they all got it wrong? Even the world renown Quad II has no issues.
If you are worried about stripping the cathodes at switch on, use a soft start controlled thyristor circuit in the the path instead of the 5Y3 or a relay controlled with a CR network. You can then leave the 5Y3 in there to look at.
Don't forget to centre tap the heater AC to reduce the hum from the 12AX7's.
So, this rectifier MUST be indirectly heated, or else it starts conducting too early, defeating its purpose."
Protect? Why does every commercial and very successful manufacturer not use anything like you suggest in your schematic? Have they all got it wrong? Even the world renown Quad II has no issues.
If you are worried about stripping the cathodes at switch on, use a soft start controlled thyristor circuit in the the path instead of the 5Y3 or a relay controlled with a CR network. You can then leave the 5Y3 in there to look at.
Don't forget to centre tap the heater AC to reduce the hum from the 12AX7's.
Let's take a look at the 5Y3GT datasheet. For a plate voltage of 380v (I suspect the voltages are higher than what the schematic says), the minimum plate series R should be 75 ohms, per plate. Here the plates are paralleled, so the datasheet recommends at least 150 ohms between the rectifier and the next capacitor (C9), which should be small. But here they used only 100 ohms and a big capacitor.
Another problem: the total current draw of the amp is around 170mA DC. At the voltages used here, the datasheet recommends less than 120mA through the rectifier.
Attachments
Because they are happy to sell you new tubes!
It's no longer the 60's, when tubes were plentiful and cheap. Today, good NOS tubes are scarce and expensive, and they should be protected.
Yes, tubes do live longer if they are allowed to heat up before applying full B+.
Last edited:
Chinese 5Y3 = 5Z2P, right ? as I can see this is a directly heated tube.
http://data.zuoshi.com/audiodata/cn/5Z2P.png
This amp is wired for 230V primary (line voltage is between 222 and 228V here).
Some infos:
PT: PW150AB-230 Power Transformer (6P6P 6P14 6P1 6P3P KT66)_E/I Core_Power Transformer_Analog Metric - DIY Audio Kit
OPT: OP9K10AB 10W 9K:4-8 Ohms Push-Pull Transformer_Push Pull_Output Transformer_Analog Metric - DIY Audio Kit
And some pictures now.
Attachments
This has nothing to do with whether the tube is directly or indirectly heated; either will work.
It is a simple matter of too much inrush current to the second cap (C9), causing flashing. Either reduce C9, increase R16 by a few times, or take the tube out of the circuit completely. Your choice.
Suggestion: Simulate the circuit to find the worst-case inrush current (ie with the tube already heated, like you flipped the power switch from ON to OFF to ON fairly quickly) and play with the component values until the current drops below the max ratings of the tube. Don't waste you time with half-assed hand waving; this is what computers were made for!
It is a simple matter of too much inrush current to the second cap (C9), causing flashing. Either reduce C9, increase R16 by a few times, or take the tube out of the circuit completely. Your choice.
Suggestion: Simulate the circuit to find the worst-case inrush current (ie with the tube already heated, like you flipped the power switch from ON to OFF to ON fairly quickly) and play with the component values until the current drops below the max ratings of the tube. Don't waste you time with half-assed hand waving; this is what computers were made for!
Chinese and Russian 5Y3 are indirectly heated, see:
5Y3 - Wikipedia, the free encyclopedia and
eurotubes-Whats-New-History
Of course they will work, only the directly heated tube will start conducting faster!
The photos provide some good news. There's room inside the chassis. The tube sockets are not mounted on the PCB. It looks like the PCB is not particularly dense and that allows for "surgery". The OP has found room for a Hammond 156R, but mounting on the side wall might be better. The idea is for the lam stack to be perpendicular to the lam stacks of the other magnetics.
From the size of the solder spots on the PCB, I'm thinking the big 'lytics are of the snap-in variety.
I've uploaded a crude drawing of a revised B+ PSU. Have at it. Forward drop in 5Y3s is large and forward drop in GZ34s is small. My thinking is that a fairly small cap. at the CLC filter's I/P will be enough to keep the rail voltage up.
From the size of the solder spots on the PCB, I'm thinking the big 'lytics are of the snap-in variety.
I've uploaded a crude drawing of a revised B+ PSU. Have at it. Forward drop in 5Y3s is large and forward drop in GZ34s is small. My thinking is that a fairly small cap. at the CLC filter's I/P will be enough to keep the rail voltage up.
Attachments
So, goodbye HT delay ?
I can try this but I guess the PS will need to be redesigned anyway, by just removing C8, hard inrush at startup is still there.
The more I think about this, the more I am confused. I asked a person who owns this amp (same PS), the 5Y3 has about 3000 hours of use and never flashed. Very surprising ?!
Last edited:
So, goodbye HT delay ?
No, not really. A much reduced HT until the 5Y3 warms up.
And without C8 how would you get a sudden inrush of current?
Hi,
Thanks for the trick.
I must have missed something, I tought charging a 270uF trough only 100R was hard for the 5Y3GT.
This morning I did some measures again, you guess what ? there is NO B+ delay !!! more than 360V on EL84 cold plates.. I don't like this.
5Z2P (Chinese 5Y3) indirectly heated ? weird.
Thanks for the trick.
I must have missed something, I tought charging a 270uF trough only 100R was hard for the 5Y3GT.
This morning I did some measures again, you guess what ? there is NO B+ delay !!! more than 360V on EL84 cold plates.. I don't like this.
5Z2P (Chinese 5Y3) indirectly heated ? weird.
Point taken, my Quad had to be re-valved when it was 25years old, as the output valves were no longer balanced. So, control the cathode current. Zero for 30seconds by disconnecting the HT with a thyristor feeding the first capacitor then gradually switch the thyristor on until full HT is set. Think about an automatic light dimmer ... or power up the screen grids when the valves have warmed up, thereby stopping anode current from flowing until you are ready and no pop at power up.
Since you have a valve recto in there, there definitely will be a soft start, it just seems fast on a human scale (1-2 seconds). But electronic components are not like humans. Receiving valves do not need a long soft start, despite what you may have read in the silly audio magazines- the EL84s will be perfectly fine and their lifetime will not be affected at all.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.