I picked up a Mr. Liang 845 SET amp in "FOR PARTS OR REPAIR" condition. It's has several no-big-deal issues and at least one "big-deal" issue. One of the floating filament supplies to the 845 triode is conducting to ground through what seems to be a failure within the circuit card. This failure extends to the 300B filament circuit and blew up some regulators too.
This amp came to me with the 300Bs but with no 845s. Given that I've got to get pretty creative to fix this thing, knowing that a good bit of the power supply card is shot, I can't see why I must stay with the 845s. Since I must replace the filament supply, and there isn't that much spare room under the hood, an external PS may be the best solution and I can build a 20 volt supply as easy as a 10. (I will note here that the other working supply is putting up 12.5 volts so it's not like I don't have to deal with that too)
I'm not going to shell out a couple of grand for a pair of NOS RCA 845s (Not that there's any thing wrong with that) and I'm not really that crazy about paying big bucks for the Chinese 845 given the mixed reviews either.
So I'm interested in some input on the quality of the Russian GM70s. Are they consistent? Are they robust? Has anyone here compared their sound to the 845?
Dave
This amp came to me with the 300Bs but with no 845s. Given that I've got to get pretty creative to fix this thing, knowing that a good bit of the power supply card is shot, I can't see why I must stay with the 845s. Since I must replace the filament supply, and there isn't that much spare room under the hood, an external PS may be the best solution and I can build a 20 volt supply as easy as a 10. (I will note here that the other working supply is putting up 12.5 volts so it's not like I don't have to deal with that too)
I'm not going to shell out a couple of grand for a pair of NOS RCA 845s (Not that there's any thing wrong with that) and I'm not really that crazy about paying big bucks for the Chinese 845 given the mixed reviews either.
So I'm interested in some input on the quality of the Russian GM70s. Are they consistent? Are they robust? Has anyone here compared their sound to the 845?
Dave
Last edited:
I've not compared the 845 and GM70 directly, but am familiar with both. I'm running GM70s in my system and am very pleased with the performance - IMHO (YMMV) I've never heard such effortless performance from another dht.. (something about fairly low rp and lots of current, not to mention heat.. )
I've run graphite plates but am currently running copper plate GM70 - I've had no reliability problems with any of the tubes I have on hand, and have heard no talk of problems with this type at all.
I've run graphite plates but am currently running copper plate GM70 - I've had no reliability problems with any of the tubes I have on hand, and have heard no talk of problems with this type at all.
not to mention heat.. )
Yes, let's not mention the heat. I live in Phoenix and I'm aware the GM70 requires twice the filament power of the 845. I may have to set it outside in the summer
Kevin, Didn't I see that you had a Power One power supply in the swap meet back in January? Do you still have that?
Last edited:
I don't know if there is any way one of tomchr's switching filament supplies could work here. I think the device he chose was a 3A device. Maybe there is a device that could eke out a bit more? Probably worth considering on a monster like this. It might cool things down a bit.
BTW, here is a link to the Mr. Liang site in case anyone is wondering what I'm going on about. 845 Amp
In the second paragraph I'm going to complain about the filament PS design, so let me say in the first paragraph that these amps are really pretty impressive in most respects. They seem to have very substantial iron, the chassis is very robust and the circuit cards, despite a failure (a cap blew a hole in it) are of very high quality. The 300B filament supply is regulated and quite satisfactory.
My big complaint is that the 845 filament supply is inadequate. The filament voltage (under load) is 12.5 rather than 10. The design is bridge rectification followed by 4 x 10,000uf caps. I have seen photos of a circuit card with a couple of additional heat sinks that make me wonder if I don't have a transitional unit assembled from a transformer wound for a regulated power supply but with the old unregulated power supply board. And yes, it is a little more hum than one would like.
I'd love to see the circuit design of a suitable regulated supply. I suppose it would need to be a constant current circuit to stand up to the inrush from the cold filament.
In the second paragraph I'm going to complain about the filament PS design, so let me say in the first paragraph that these amps are really pretty impressive in most respects. They seem to have very substantial iron, the chassis is very robust and the circuit cards, despite a failure (a cap blew a hole in it) are of very high quality. The 300B filament supply is regulated and quite satisfactory.
My big complaint is that the 845 filament supply is inadequate. The filament voltage (under load) is 12.5 rather than 10. The design is bridge rectification followed by 4 x 10,000uf caps. I have seen photos of a circuit card with a couple of additional heat sinks that make me wonder if I don't have a transitional unit assembled from a transformer wound for a regulated power supply but with the old unregulated power supply board. And yes, it is a little more hum than one would like.
I'd love to see the circuit design of a suitable regulated supply. I suppose it would need to be a constant current circuit to stand up to the inrush from the cold filament.
Last edited:
tomchr's is a switching regulator (constant voltage) but with a soft start. You can bring the filament up as slow as you want.
However, he definitely did not have a tube like this in mind when he designed it. It may not handle the voltage or current. It's worth looking into, though.
However, he definitely did not have a tube like this in mind when he designed it. It may not handle the voltage or current. It's worth looking into, though.
Kevin, Didn't I see that you had a Power One power supply in the swap meet back in January? Do you still have that?
As I commented tonight in swap meet I may yet have those supplies - I'll check in the morning as I head out to work and let you know. Hopefully you are local as I had hoped to hand them directly to a new owner.
I'm pretty familiar with the Liang 845 and there do seem to be several versions - they aren't very common here in the U.S.
I'd love to see the circuit design of a suitable regulated supply. I suppose it would need to be a constant current circuit to stand up to the inrush from the cold filament.
Dave, here's a GM-70 filament regulator already in service in dozens of GM-70 amps around the world. Similar versions are supplying 813s [5A] and even 75TLs [6.25A].
Yes, it is current-driven - this also prevents filament breakage from turn-ON surge for susceptible filaments (Svetlanas, and many modern tubes).
It is designed to get the best possible sound from filament heating, & plenty of DIYers like it:
http://www.diyaudio.com/forums/tubes-valves/38248-new-dht-heater-9.html#post2794791
A self-assembly kit - PCB + parts, please send email/PM if you'd like the PDF app notes or more details.
Attachments
Hopefully you are local as I had hoped to hand them directly to a new owner.
Well I'm not. But there is an electronics surplus store here in town with a big pallet of Power One power supplies, and I'll probably need two of them anyway, so no worries.
Bummer... I was hoping to palm them off, err give them away to anyone who could use them.
I found an Antek toroid, bridge large cap and LT1084 regulator to be a better match to Rod's filament CCS than the Power Ones.. Note however that the LT1084 will go into current limiting with this load without a little help so I placed a 8.2 ohm resistor (25W) in parallel with the regulator to help it start, from a cold start this provides roughly 3A, and once the regulator comes out of current limit this resistor provides 0.5A of the overall load current, and dissipates just a couple of watts. The low output Z of the regulator dominates from an AC perspective once the regulator is running and the ripple level is just a few mV.. (One potential issue with this approach is if a filament goes open the voltage would soar to the unloaded supply voltage of the transformer used - depending on choice of transformer I might recommend a voltage clamp set to a level safe for the Coleman CCS be used.
Running at 3A those PO regulator's power transformers can be pretty noisy. They are good for experimenting and powering stuff short term..
I found an Antek toroid, bridge large cap and LT1084 regulator to be a better match to Rod's filament CCS than the Power Ones.. Note however that the LT1084 will go into current limiting with this load without a little help so I placed a 8.2 ohm resistor (25W) in parallel with the regulator to help it start, from a cold start this provides roughly 3A, and once the regulator comes out of current limit this resistor provides 0.5A of the overall load current, and dissipates just a couple of watts. The low output Z of the regulator dominates from an AC perspective once the regulator is running and the ripple level is just a few mV.. (One potential issue with this approach is if a filament goes open the voltage would soar to the unloaded supply voltage of the transformer used - depending on choice of transformer I might recommend a voltage clamp set to a level safe for the Coleman CCS be used.
Running at 3A those PO regulator's power transformers can be pretty noisy. They are good for experimenting and powering stuff short term..
Hi Kevin, It's OK, for 845/GM-70 regulators, the no-load voltage can go to 35V without anything bad happening. If that's not high enough, you can switch the little cap C1 for a 50V rated part, and 50V unloaded will be tolerated. For anyone needing higher, the signal FETs can be switched, and 100V is acceptable! Even 400V, with uprated power transistors. Discrete design has many advantages!
Under load, 25v to 29v is OK, but staying at the lower end keeps the heatsink modest.
Under load, 25v to 29v is OK, but staying at the lower end keeps the heatsink modest.
Also consider you can swap in that amp a triode strapped 813 which is easy to find, cheap, sturdy and ... yes it sounds great too.
I looked at that but the 5 amp filament requirement presents a problem for the 3.25 amp transfo windings. That and the socket change are negatives. Price and availability are positives.
But I'm only getting 11.4 vac from the existing windings, so they aren't quite adequate for use with the Coleman regulator anyway. What I really need is an 8 volt triode.
Bummer... I was hoping to palm them off, err give them away to anyone who could use them. <snip>
You are using an LT 1084 on a 20V GM70? I didn't realize they were available above 12 or 15 volts. OK, I just checked the data sheet and TIs adjustable ones top out at 15 volts. Which one are you using?
And thanks for the word on the PowerOne supplies.
Last edited:
Hi,I've not compared the 845 and GM70 directly, but am familiar with both. <snip>
Do you think GM70-graphite plate can made a better sound quality than the 6C33??
Thanks
I looked at that but the 5 amp filament requirement presents a problem for the 3.25 amp transfo windings. <snip>
You can use same socket for 813 and GM70 (just drill the holes a bit), I'd go for a 813/GM70 amp and toss a little more money on the filament trafos (money you saved buying those bottles vs 845).
8V tubes ... hum ... what about a 801A? you need to rework the PSU (it only takes 500V on the plate) and maybe your OPT is not the best you can use but 801A sounds great too. I don't know anything about your amp but I guess you are running at something like 900V on top of the 845 so maybe there is room for a direct coupled 300B-801A stage.
Eeek! I hope you paid pennies on the pound for that broken Chinese amp!
At this point, why not just implement an outboard 10V power supply for filaments? Put
a common mode choke in that 845 chassis though! There's some nice Triad/Magnetek
4 Amp CMC's out there - check the Allied Catalog.
The next catastrophe will be the HV supply.. Something will go... Arc out from dust or
corona breakdown. Watch those balancing / divider resistors on series electrolytic caps too!
At this point, why not just implement an outboard 10V power supply for filaments? Put
a common mode choke in that 845 chassis though! There's some nice Triad/Magnetek
4 Amp CMC's out there - check the Allied Catalog.
The next catastrophe will be the HV supply.. Something will go... Arc out from dust or
corona breakdown. Watch those balancing / divider resistors on series electrolytic caps too!
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
- Home
- Amplifiers
- Tubes / Valves
- 845 vs GM70