Hi guys,
Please, what is the best GM70 tube version suited for a SET parallel amp??
I know three versions of this tube:
Copper plate(Ulyanovsk),
Graphite plate(Ulyanovsk)
Metal Plate(Photon Uzbekistan plant).
Anyone know what is the output plate impedance of this tube??
The PDF specs files I have see do not show this detail.
Thanks for any info, Gustavo
Please, what is the best GM70 tube version suited for a SET parallel amp??
I know three versions of this tube:
Copper plate(Ulyanovsk),
Graphite plate(Ulyanovsk)
Metal Plate(Photon Uzbekistan plant).
Anyone know what is the output plate impedance of this tube??
The PDF specs files I have see do not show this detail.
Thanks for any info, Gustavo
Many people will tell you the copper plate version is best (probably because it is the most scarce 😉 )
Technically the graphite plate should be superior, it can radiate heat more efficiently and can stand higher temperatures.
My suggestion would be (and some people will disagree with this), to use the cheaper graphite types and spend the saved money on better output transformers.
You can calculate rp as rp=mu/transconductance. However, for the GM70, the parameters vary a lot with B+. Most schematics I have seen use quite high load impedances, such as 10 kohms, for high B+ like 1000--1200V.
Kenneth
Technically the graphite plate should be superior, it can radiate heat more efficiently and can stand higher temperatures.
My suggestion would be (and some people will disagree with this), to use the cheaper graphite types and spend the saved money on better output transformers.
You can calculate rp as rp=mu/transconductance. However, for the GM70, the parameters vary a lot with B+. Most schematics I have seen use quite high load impedances, such as 10 kohms, for high B+ like 1000--1200V.
Kenneth
To my point of view copper plate is better sounding, but graphite plate version is more powerfull.
I am building a 12HL7/GM70 and have the graphite tubes for now. I have friends that have the copper plate version. Once the amp is done and broke in nice, I will roll the coppers in it and see how they sound. Do you have a scheem yet? What B+ are you running? two stage or three stage?
I'm also in the process of building a GM70 amplifier and for the time being will use graphite plate GM70, but I am tentatively thinking of getting some copper plate ones later. I'm running this design pretty hot so for copper I would probably have to reduce the plate current significantly.
Others more familiar with the GM70 than I currently am have indicated that both types have their virtues, but that possibly the copper version sounds slightly better. So far I have heard only the graphite type and it sounded just fine to me..
I'm going to be running 1KV/125mA with graphite, driven by D3A via IT. Both stages will be fixed biased, and supplies fully regulated. (On separate chassis)
Project is moving rather slowly, I have posted details on another long dormant thread. (When I was planning to use choke loading.) Once I actually start putting the thing together I will post to that thread.
Others more familiar with the GM70 than I currently am have indicated that both types have their virtues, but that possibly the copper version sounds slightly better. So far I have heard only the graphite type and it sounded just fine to me..
I'm going to be running 1KV/125mA with graphite, driven by D3A via IT. Both stages will be fixed biased, and supplies fully regulated. (On separate chassis)
Project is moving rather slowly, I have posted details on another long dormant thread. (When I was planning to use choke loading.) Once I actually start putting the thing together I will post to that thread.
Hi FRM
I find the graphite more detailed and sharp or quick and the copper maybe has a tad deeper bass but not as much power. I have a pair of the extreamly rare plain metal plate with brass base but haven´t got round to hear them out yet, one of these days.......
I find the graphite more detailed and sharp or quick and the copper maybe has a tad deeper bass but not as much power. I have a pair of the extreamly rare plain metal plate with brass base but haven´t got round to hear them out yet, one of these days.......
Hey Kevin, I am running 840v B+ at somewhere around 80 to 100 ma. My horns are 115db efficient so I don't need more that the 7 to 10 wpc I will get out of it. I am also using the OA2 controll tube on the 12HL7.
Hi Boys,
Thanks for your valuable and detailed info. Iam surprised how many GM70 amps are being manufactured.
This GM70 looks a good tube for the price, in view of a 500dollars 845 single.
Go ahead it worth the work.
Thanks for your valuable and detailed info. Iam surprised how many GM70 amps are being manufactured.
This GM70 looks a good tube for the price, in view of a 500dollars 845 single.
Go ahead it worth the work.
Mr.Kenneth: Thanks for this formula. So, as you said, looks this GM70 has a plate impedance of 9k ohms!! Are this too high? This could be a problem?
As a 845 plate impedance is around 3k ohms. Looks the output transformer will be big...
Regards, Gustavo
Dear H00HBT,
By this good description you and others said on GM70 sound quality, I will choose the inexpensive Graphite Plate.
Iam in a very begining stage of choose the parts, buy it and store for later building. I will no build myself, as I know only made solder, I will use a tube builder in my city, or any other in Brazil that accept the task.
Many thanks for your support guys, Gustavo
I have 8K output transformers. If I went to 5K that would work just fine and give me more power output but with higher distortion. The 8Ks will give me all the power I need and much less distortion. Cheers.
Hi Gustavo,
the values for mu and transconductance in the datasheet have such a high tolerance on them, that it's hard to calculate the exact rp. The best way would be to determine it graphically from the curves.
Take a look at the Ia/Ua curves in this datasheet: http://frank.pocnet.net/sheets/084/g/GM70.pdf
You can graphically determine rp by measuring the slope of the curve, in the working point (bias point) of your tube. In most cases you will find something around 2000 ohms.
With a triode, you should always use a load impedance somewhere between 2 to 5 times the rp. This is why you typically see GM70 with output transformers with 6 ... 10 kohms primaries.
As DJN says, if you use low load impedances, such as 2x rp, you get high power but high distortion. If you use a load impedance of, say, 5x rp, you get lower power but much less distortion. It is up to you, to decide which you want 🙂
I read in your first post that you want to use two tubes in parallel. If you do that, you make one big virtual tube with the same mu but half the rp of a single tube. So your OPT primary impedance must be halved, too. And you need to take precautions to insure that both tubes will share the load 50/50.
Hope this helps,
Kenneth
Hello Mr.Kennteh,
Yes this help me alot, many thanks for your help.
I need 50W 8 ohms from this two GM70 tubes, so as I see in other thread here about the GM70, seems the B+ will be 1000 to 1100Volts, with a xx Bias value(maybe 125mA as I just read in an DJN reply).
What is the best input stage tube for the GM70??
I see some builders here are using the pentode 12HL7, they use as Pentode or wired as Triode. I would like a Triode in the input stage, as 6SN7 or 12AX7 family, or any other good Triode.
Do is need use two tubes in the input stage??
If yes, I will use the Mu Follower circuit for input stage.
Thanks again for any tip, Gustavo
Last edited:
"Mr." is not necessary 🙂
With B+ of 1000 to 1100V, the GM70 is relatively easy to drive, even using RC coupling is a possibility. However, you need a lot of voltage swing to drive it to the fullest. This is why many people use a pentode with a high amplification factor. 12GN7, D3A, 6C45, ... Ideally you want types with sufficiently high plate voltage rating (400V) to get the required voltage output swing, and sufficient transconductance to allow good input sensitivity.
If you use triodes you need two input stages to get enough voltage gain, unless your preamp has a very high level output. Maybe 6SL7 in first stage, followed by something having low rp and which can stand a 400V plate voltage, could work.
Another possibility is an interstage transformer with, for example, 1:2 step up ratio which can then be driven by a tube with a lower output voltage swing. Triodes such as the 2A3 or 300B then become a possibility. You can find many schematics on the net that work like this. But a good IT is not cheap... 🙁
As you see, lots of choices to consider 🙄
Another thing... if you can make a very good input stage you could get 35...40W from a single GM70. The difference in loudness between 40W and 50W is not that big. Maybe you could start with a single tube and then decide if you need two in parallel. Parallel SE is more hassle to get to work properly.
Kenneth
With B+ of 1000 to 1100V, the GM70 is relatively easy to drive, even using RC coupling is a possibility. However, you need a lot of voltage swing to drive it to the fullest. This is why many people use a pentode with a high amplification factor. 12GN7, D3A, 6C45, ... Ideally you want types with sufficiently high plate voltage rating (400V) to get the required voltage output swing, and sufficient transconductance to allow good input sensitivity.
If you use triodes you need two input stages to get enough voltage gain, unless your preamp has a very high level output. Maybe 6SL7 in first stage, followed by something having low rp and which can stand a 400V plate voltage, could work.
Another possibility is an interstage transformer with, for example, 1:2 step up ratio which can then be driven by a tube with a lower output voltage swing. Triodes such as the 2A3 or 300B then become a possibility. You can find many schematics on the net that work like this. But a good IT is not cheap... 🙁
As you see, lots of choices to consider 🙄
Another thing... if you can make a very good input stage you could get 35...40W from a single GM70. The difference in loudness between 40W and 50W is not that big. Maybe you could start with a single tube and then decide if you need two in parallel. Parallel SE is more hassle to get to work properly.
Kenneth
I bought a lot of what I have from a friend that was collecting parts for the build. I got all the iron needed and the GM70 tubes plus sockets. The OTs were custom wound by a company that is no longer in business.
Hi Kenneth, yep, I went back and forth on all this. I ended up with going with the 12HL7 with a O2A control tube as well. My speakers are 115db eff. so the 7 to 10 wpc I will get is plenty. Currently I am running my speakers with a Decware Zen amp putting out 1.5wpc and I never turn it up more than 1/3 the way.
I am going to draw my scheem this weekend and will post it. Cheers.
p.s., my B+ is 840v
I'm using 7K:8 transformers in mine which were conservatively designed based on several load line scenarios at 5K, 6K, and 7K - 5K provided both the highest output power, highest distortion, and lowest damping factor. I don't remember the exact rp I calculated at the time, but IIRC for my chosen operating conditions it was significantly less than 2K. Calculated output power steering clear of class A2 with the 7K primary was about 20Wrms. I wanted low distortion and a reasonably low source impedance and don't really need that much power, linearity is more of a concern.
I ordered a pair of Lundahl IT for the driver stage today, (5wk lead time for the model I ordered as it is non-stocked) some limited class A2 operation is now possible. (At least a few mA of grid current) I'm using the D3A in triode connection to drive the GM70 - based on bench measurements this single stage driver will swing >300Vpp on a 250V supply and mu is almost 80. Originally I had planned on LED bias, but will go with fixed bias instead to minimize rp.. (The dynamic impedance of the LED I was using is about 10 - 12 ohms, which when multiplied by mu + 1 gives me way more rp than I can afford with the IT, it was acceptable with a 100H choke load.)
Output stage running at 1kV and ~125mA.
Last edited:
Dear Kenneth,
Your posts are striking instructive
thanks for it
Iam impressed with the tubes subject, it is not a simple or easy matter.😱I definitely do not want use a interstage Transformer, I prefer use a Pentode then.
As you said two GM70 in SEP made one big virtual tube(I did not like this tube scenario/behaviour) Iam chaging my project to two GM70 in Ciclotron, maybe a bit more expensive than SEP, but a more reliable topology in my newbie view.
What is ''RC Coupling''?? (maybe Resistane Cathode coupling?)
Do you see any problem with fixed Bias (B+ 1kv and 125mA) with two GM70 in ciclotron??
I want the least possible parts, so just one tube in the input stage is great, seems good a low price and abundant tube on the market(not sure what tube meet this requeriments) and yet have good sound.
Maybe the Russian 6C45, to stay only with russian tubes.
I use a fullrange speaker now, but may change in the future to a Klipsch RF83, specified 100dB/SPL, 8ohms(but minimum 3.6ohms at 179Hz) and two-way cut at 1650Hz http://www.klipsch.com/na-en/products/rf-83-overview/
Thanks for advicing, Gustavo
Last edited:
- Status
- Not open for further replies.