Hoffman Amplifiers Tube Amplifier Forum
Amp Stuff => Tube Amp Building - Tweaks - Repairs => Topic started by: Rev D on June 26, 2010, 02:26:03 pm
-
I've decided to add a pentode on the regular channel of my 18 watt TMB amp. The regular channel I just don't care for and rarely use, I have the layout I want to use to add the pentode on that channel which is basically a volume and 6 pos tone switch like the Matchless I believe.
EF86's are sort of expensive and known for being microphonic, there is the 5879 that many of you have used for the Hoso's etc which are'nt to bad at about 7-9 bucks a piece pretty reasonable. Then I find the EF184 (phillips) are crazy cheap at a tad over a buck. I've seen posts on the net where someone used one in a ac15 and liked them just fine. I know they'll work, but my question is how do they fit in terms of gain vs the EF86 and the 5879? For the money I'm definitely willing to experiment with it and play with the circuit to get it to work, but has anyone had any first hand experience with them? Thanks for any info you may have in advance.
Regards,
D.
-
Never tried them, don't know anything about them. So I went to Frank's and looked up data sheets for the EF86, 5879 and EF184.
The basic figure of merit for pentodes is Gm. That is, how many mA/volt (to correlate to data sheets, 1000 micromhos = 1 mA/v). The reason is due to the high internal resistance of pentodes, the easiest way to figure gain for a pentode is Gm * Load Resistance (Rl).
The problem is, data sheets don't always give you an apples-to-apples comparison of tube characteristics. But they can help you guess a good bit.
The 5879 has a Gm of 1.2-1.5 mA/v at somewhat reasonable plate current. The EF86 has a Gm of 2 mA/v at a similar plate current. You'd expect a little more gain from the EF86.
The EF184 has a Gm of 15mA/v (!) but at a plate current of 10mA. That's a lot of current to pass, and fortunately the data sheet showed how Gm drops off at lower current. It's fine until you get down around typical guitar amp ranges, then Gm is around 4mA/v.
That's still a lot of gain. Assuming screen voltage and bias voltage lands each tube to where they have these predicted Gm values, compare what gain you get with a 100k resistor for each.
5879: 1.2mA/v * 100k = 120
EF86: 2mA/v * 100k = 200
EF184: 4mA/v * 100k = 400, but at much higher stage current.
These numbers are representative only. We haven't figured for the loss of gain due to the following stage grid resistor, and real stage gains are also lower due to lower stage current and therefore lower Gm. But you could probably expect double the gain from an EF184. To make use of that, you'd have to seriously look at how big an input signal you'll be applying, and how much supply voltage is needed to allow the tube to make an output that is actually 400 times the input. You might find you can't get there from here.
-
Thanks for your time, much appreciated. Wow I never expected it to be higher gain than the ef86. I may order a few to play with and get a 5879 as well since I know people have had good luck with them. If I'm able to accomplish anything with it I'll report it. Once again thanks for the time, much appreciated.
Regards,
D.
Ps: If I can do nothing on the 18watter with the ef184 there is a plan for a Moonlight amp on ax84 I may investigate, I could use a small wattage amp in the apt.
-
EF184, similar to a 6EJ7, a sharp-cutoff pentode used as a TV IF amp. Explains why it's so cheap, especially with that level of gain and current.
-
I have used an EF184 for experiments but did not test it more thoroughly because it was very microphonic. I also tried various other cheap pentodes (e.g. EF85, ECF802, ECF80) but found them all too microphonic to be useable.
-
Yeah, thats what I've heard, I picked up a few just for grins, I've heard tail that the ones with the mesh inside are less microphonic (and hence more expensive) these being cheap I'm sure they'll be microphonic, but I may play with them all the same to see what I can do with them. Thanks for your views on the tube.
Regards,
D.