AI preamps are notorious tube "eaters", in the stock configuration. The ONLY tube to use in a stock configuration AI preamp is the Russian 6H23П-EB (6n23p-ev). Tubes marketed as 6922, under the ElectroHarmonix (EH) label are (in fact) 6H23П-EB.
Check the archives here and over on AA for circuitry changes that allow other 6922 varieties to survive.
Check the archives here and over on AA for circuitry changes that allow other 6922 varieties to survive.
That is very interesting, and news to me (I use EH 6922 in my RTP3). I haven't been able find a datasheet for the EH 6922. What parameters are superior to the standard 6922?
Alex
The M3A has an extremely complex power supply, with monolithic regulator (LM723) feeding a Mosfet device. It is not so effective and has a highish output impedance, which is not desired for a regulator.
In addition, the rail voltage is way too high causing tube stress as well as highish gain, so the preamp produces a significant level of noise.
My suggestion:
1. Replace the on-board regulators with something like the Salas high-voltage regulator - one for the line stage, another one for the phono stage.
2. Reduce the rail voltage to about 110V - it will reduce the tube stress and reduce the gain a bit.
3. Lose the cathode bypass cap - this will significantly reduce the gain as well as background noise.
4. The plate resistor in the M3A is made of 3 parallel resistors of 12K each, which means 4K. You may take 1 or 2 out, and further reduce tube bias and gain.
Good luck !
Yair
In addition, the rail voltage is way too high causing tube stress as well as highish gain, so the preamp produces a significant level of noise.
My suggestion:
1. Replace the on-board regulators with something like the Salas high-voltage regulator - one for the line stage, another one for the phono stage.
2. Reduce the rail voltage to about 110V - it will reduce the tube stress and reduce the gain a bit.
3. Lose the cathode bypass cap - this will significantly reduce the gain as well as background noise.
4. The plate resistor in the M3A is made of 3 parallel resistors of 12K each, which means 4K. You may take 1 or 2 out, and further reduce tube bias and gain.
Good luck !
Yair
I'm new here and stumbled on this thread. I recently bought a 1996 (1st production year) M3A on eBay an it looks impeccable. Regarding tube life: In reading the manual there are strong indications that when turned OFF the unit is in a pre-warmed state and that the delay before it's ready is there for warm-up. At least that's the interpretation I got. The reason given was to get around the thermal shock the filaments take upon normal cold turn-on. Good idea I thought. Well it's just not true. The filaments NEVER turn off and I thought this was a fault in my unit so I emailed AI. I got to speak with Art Ferris (designer and owner). There are a number of things I learned in that conversation that I thought I'd pass on.
1) No we don't pre-heat the tubes - causes cathode contamination.
2) Through the many years we have had to search for batches of acceptable tubes in the 6DJ8/ECC88/6922/E88CC family. They have to discard many.
3) The line-stage is particularly picky. Those two tubes (V3 and V4) are run in parallel and have to be highly matched pairs meaning each half must match the other two halves (i.e. ya just can't roll these at will). All four plate voltages should land at 125Vdc + or - 5 to 6V.
4) The phono stage is less picky.
5) The best tubes they like now are the Russian 6N23P. They have used 6922/E88CC in the past. If they'd found non-microphonic 6DJ8s they would have used those.
6) I was told by Art to set the gain switch to HI and the added bypass capacitor switch to OFF.
7) Their heater filament supply does not use a regulator. They use a PI filter with a series 10 ohm current-limiting R - one for each stage. This tid-bit came out when I asked about the use the Russian 6N1P which is a family member BUT the heater current is 2x that of the 6DJ8/6922 @600mA vs. 300mA. Since the heater supply is NOT regulated, using these will drag down the filament voltage and could over-heat the 10 ohm power resistor. So, if you were to try the very-good-in-my-opinion 6N1P then you must halve that 10 ohms and the easiest way is to parallel another 10 ohm across the original. Use the same wattage and make sure you are doing the mod to the stage using the two 6N1Ps. Art said the power supply is beefy so one can do this.
So to make a long story short - the heaters run full-on, 24/7/365. That's just shy of 9,000 hours! Aren't even long-life military tubes rated for a fraction of that time? And, if the four tubes are lit then what is this 30 second "warm-up" timer for. The ON/OFF button is RED, you press it to ON and the preamp waits 30 seconds to become active - why? For what reason? It's just an engineering curiosity of mine (b/c I am an EE) and I forgot to ask.
There is no regulator so the filament voltage does not snap to 6.3V upon turn-on. There are 10 ohm series resistors which absorb the inrush current. The voltage/current rises rather slowly. It's not like a light bulb. I use the switch on my power conditioner to turn the M3A OFF.
1) No we don't pre-heat the tubes - causes cathode contamination.
2) Through the many years we have had to search for batches of acceptable tubes in the 6DJ8/ECC88/6922/E88CC family. They have to discard many.
3) The line-stage is particularly picky. Those two tubes (V3 and V4) are run in parallel and have to be highly matched pairs meaning each half must match the other two halves (i.e. ya just can't roll these at will). All four plate voltages should land at 125Vdc + or - 5 to 6V.
4) The phono stage is less picky.
5) The best tubes they like now are the Russian 6N23P. They have used 6922/E88CC in the past. If they'd found non-microphonic 6DJ8s they would have used those.
6) I was told by Art to set the gain switch to HI and the added bypass capacitor switch to OFF.
7) Their heater filament supply does not use a regulator. They use a PI filter with a series 10 ohm current-limiting R - one for each stage. This tid-bit came out when I asked about the use the Russian 6N1P which is a family member BUT the heater current is 2x that of the 6DJ8/6922 @600mA vs. 300mA. Since the heater supply is NOT regulated, using these will drag down the filament voltage and could over-heat the 10 ohm power resistor. So, if you were to try the very-good-in-my-opinion 6N1P then you must halve that 10 ohms and the easiest way is to parallel another 10 ohm across the original. Use the same wattage and make sure you are doing the mod to the stage using the two 6N1Ps. Art said the power supply is beefy so one can do this.
So to make a long story short - the heaters run full-on, 24/7/365. That's just shy of 9,000 hours! Aren't even long-life military tubes rated for a fraction of that time? And, if the four tubes are lit then what is this 30 second "warm-up" timer for. The ON/OFF button is RED, you press it to ON and the preamp waits 30 seconds to become active - why? For what reason? It's just an engineering curiosity of mine (b/c I am an EE) and I forgot to ask.
There is no regulator so the filament voltage does not snap to 6.3V upon turn-on. There are 10 ohm series resistors which absorb the inrush current. The voltage/current rises rather slowly. It's not like a light bulb. I use the switch on my power conditioner to turn the M3A OFF.
A race engine requires high octain gas or it detonates and cracks pistons and the AI requires the best tubes, not me thinks I,ll put "this" set in and give it a spin.,
Looking at an example bell curve, with this many preamps out there, and there are many!
You have that very small group that likes to roll their own and then complain at the same time.
If time is money to you, then buy AI,s set of russian tubes or you spend considerable time like I have with testing them, matching, waiting to form retest agian only to find a close match and then they are microphonic, etc.
Its hard to find a 6h23p that has it all, tight match, low noise, no microphonics. At least 1/2 of mine are not acceptable for use in this pre. It just takes time that many don't have if you want to DIY
This is a great preamp so feed it what the maker suggests and you should be happy, and if not, then sell it and move on
Looking at an example bell curve, with this many preamps out there, and there are many!
You have that very small group that likes to roll their own and then complain at the same time.
If time is money to you, then buy AI,s set of russian tubes or you spend considerable time like I have with testing them, matching, waiting to form retest agian only to find a close match and then they are microphonic, etc.
Its hard to find a 6h23p that has it all, tight match, low noise, no microphonics. At least 1/2 of mine are not acceptable for use in this pre. It just takes time that many don't have if you want to DIY
This is a great preamp so feed it what the maker suggests and you should be happy, and if not, then sell it and move on
Hi again,
Just wondering what you folks think about running Mod 3A tube filaments 24/7/365.
I have a CAT SL1 MkIII and it does not do that. It actively ramps-up the filament voltage upon turn-on and THAT is why one waits for the READY light.
Is there a Mod 3A schematic around?
Now, with my new-to-me Mod 3A there is an issue with 60Hz hum in the left Phono channel. Curious thing is that with the cover off, placing my hand over V2 (left chan phono) the hum level reduces dramatically. Same effect if I just lay the cover on the chassis w/out screws. It's almost gone. I changed both V1 and V2 with German NOS with no change. AI's Art Ferris (designer/owner) insists that I send it back to them with all the associated hassle, shipping $, insurance $, their bench fee ($85), their repair cost, and return shipment $$. I, as an EE and still-alive tube-gear owner for decades asked for some simple guidance as to what to check, measure, scope, visually look at, and so on. He will not budge. In my last email I stated the above again and offered to pay a consult fee in order to speak with their repair person. Am waiting to hear back.
Any ideas re this behavior?
Just wondering what you folks think about running Mod 3A tube filaments 24/7/365.
I have a CAT SL1 MkIII and it does not do that. It actively ramps-up the filament voltage upon turn-on and THAT is why one waits for the READY light.
Is there a Mod 3A schematic around?
Now, with my new-to-me Mod 3A there is an issue with 60Hz hum in the left Phono channel. Curious thing is that with the cover off, placing my hand over V2 (left chan phono) the hum level reduces dramatically. Same effect if I just lay the cover on the chassis w/out screws. It's almost gone. I changed both V1 and V2 with German NOS with no change. AI's Art Ferris (designer/owner) insists that I send it back to them with all the associated hassle, shipping $, insurance $, their bench fee ($85), their repair cost, and return shipment $$. I, as an EE and still-alive tube-gear owner for decades asked for some simple guidance as to what to check, measure, scope, visually look at, and so on. He will not budge. In my last email I stated the above again and offered to pay a consult fee in order to speak with their repair person. Am waiting to hear back.
Any ideas re this behavior?
Eli Duttman said:
Could there possibly be some connection between these two statements? Heaters on but no anode current is precisely the correct situation for cathode interface problems. Some valves don't suffer from this (it depends on the exact formulation of the cathode metal), but most do. The symptoms can be similar to loss of cathode emission.JRA said:
If you want small-signal valves to last a long time then there are two options:
1. use them in the 'On' configuration with about the right heater voltage and a reasonable amount of anode current - but not too much to cause overheating.
2. switch them 'Off' (no heater, no supply rail voltage) so when out of use they are doing nothing at all.
Swapping between these two safe modes is fine, as long as you don't take too long to do it.
The other thing to note is that fussiness about valves is generally a sign of poor circuit design. I am sure this particular unit is the exception which proves the rule!
If the heater supply has a fixed resistor inline with the heater's thermally-variable resistance then why keep them on for 9,000 hours/yr? There is no cold-filament shock. Period. AI is being goofy. Even the best Russian 6n23P military tubes are only rated for a lifespan of 5,000 to 10,000 hours and, at the 5,000 hour mark they make note of the expected degradations.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.