diyAudio (
-   Tubes / Valves (
-   -   1AG4 Direct Heated Pentode, Output (

TheGimp 12th February 2012 03:49 AM

1AG4 Direct Heated Pentode, Output
Anyone ever used them?

They look like they were designed for driving headphones. The data sheet shows a speaker, but the spec is 35mW at 12%thd.

That said, the Raytheon data sheet gives triode connected curves and it looks promising.

I'm thinking of using a pair in an LTP with a transformer anode load as the output for a DAC. 2 - 2.5V bias at 35V on the anode looks pretty linear.

revintage 12th February 2012 12:49 PM

Curves look really great. But mu is low, should be in the ballpark of 5.

TheGimp 12th February 2012 05:50 PM

Low Mu is OK since the DAC output is 2vrms differential. I really don't even need gain, just buffering to the transformer for the output. At that, the transformer loaded LTP may be overkill.

I was considering the WSM10K/600 which has a primary inductance of 22H and would provide a step down ratio of 4.1:1 to get the output voltage back down close to line level, and increases the anode load impedance to make the load higher to match to the tube specs.

With a pentode ra of 120K, and Mu of 5, in triode mode I should see an ra of near 24K. With the 4.1:1 transformer the load can be as low as 5.8K and match the tube ra. (rough approximation).


TheGimp 14th February 2012 08:03 PM

Start with a 7.5V wall Wart.

Feed into a 78L05, with .1uF on input, 10uf and .1uf on the output (all ceramic).

Use this to feed a MAX253 at 300KHz driving a coilcraft 044-70-CL 1:1 ct transformer. with the ct of the input feed from the wall wart. output FW rectification with FRED diodes, followed by 2 ea 10uF ceramic, one .1uF ceramic.

Output = 9V.

Post filter with 22uH inductor followed by 10uF ceramic.

Feed this into a current source, based on the TL431 with a pass transistor.

Set I out for 40mA.

This is my proposed heater supply.

Can I measure noise by using a 10uF ceramic cap and ac couple the output into my PC sound card while running AudioTester set for FFT?

Is there a better way?

thorstenlarsen 14th February 2012 08:23 PM

Hi Gimp!
How about the bass?
Is 22 H enough? Compared to innner resistance of tubes// loud.

TheGimp 14th February 2012 08:29 PM

I don't know if 22H is enough or not, so I plan on building a prototype to test it. I'll use the 10K:10K I have on hand to see if it works. If so I'll order a couple to 10K : 600R to try.

I'm trying to figure out the heater supply right now. I need to supply 4ea, 1.25V heaters with 40mA each, and they need to be isolated. I'm going to try to measure the noise on this switcher and see if it looks like it might be quiet enough with sufficient filtering.

TheGimp 15th February 2012 03:22 AM

5 Attachment(s)
I ran a couple of scans tonight to see how noisy the switcher would be.

First is everything wired up but the wallwart not plugged in.

Second is with power.

Third is with a 470uF added across the load resistor.

fourth is with an additional 0.1F Supercap across the load.

Fifth is with four AA Batteries running it.

I wonder if the noise groups that look like IM above 2K, 3K, etc were from the switcher.

They are down pretty low, but I still don't like them. I think I will rework the switcher tomorrow and add snubbers/Ferites to the rectifiers to see if they help suppress noise.

TheGimp 18th February 2012 05:51 PM

I'm taking some measurements that look promising.

However, I got a thought that may impact the results.

The data sheet shows a filament voltage of 1.25V at 40mA.

However, the data sheet also states "...designed for use in the output stage in battery operated receivers."

Now this is back in the mid-50s since my dad was flying RC airplanes by 1956. Tube data sheet shows March 1, 1955.

The batteries then were either Lead Acid (2.2V/Cell) or Carbon Zinc batteries (1.5V/Cell). The carbon zinc cell nominal voltage was 1.5V, not 1.25V. Though they do have a rapid drop in voltage under load.

So I'm going to crank the filament supply up to 1.35V instead of 1.25 and see how it operates.

DavesNotHere 18th February 2012 08:15 PM

I remember building a se amp that the 1AG4 drove a 6v6. the heater of the 1 ag4 was in the cathode of the 6v6 the 6v6 was battery power heater that also battery biased the grid.

40VDC plate supply (I used the bias tap of the 250V 6v6 plate power transformer as the plate supply )

later on I switched the battery 6v6 heater to 6VAC winding, replace motorcycle batteries for 4 AA's

TheGimp 18th February 2012 08:33 PM

2 Attachment(s)
I like to use instruments to take measurements to find the trade-off for a circuit before I bother listening to it. This circuit is a classic example of why.

I started from the data sheet which looked like it should be linear in the same bias region as recommended for pentode operation.

I built the circuit with an adjustable current source in the tail, figuring I could tweak it to minimize distortion.

What I found was that the tubes I'm using show better linearity down at 1mA, and have much higher Mu, with a measured gain of 11.69 from the plate transformer half secondary. This is a 10K:10K transformer with ct on both sides. So the gain from a half secondary should correspond to one tube.

At two mA I only got almost 20dB less output! there is a very dramatic shift in gain around 1.5mA.

The second thing I found is that the tubes have much higher gain and linearity when operated at 1.35V on the heater vs 1.25V. 1.25V operation yields about half the gain of 1.35V operation. Since these tubes were designed for battery operation, I believe this to be an acceptable operating point.

If I had built the amp prototype and went directly to listening tests, I certainly would have spent much more time trying to tune the sound compared to measuring the amp first. Now I can isolate a much smaller set of variables to play with while listening to the amp, and have a better understanding of what is going on inside the amp.

All times are GMT. The time now is 06:30 AM.

Search Engine Optimisation provided by DragonByte SEO (Pro) - vBulletin Mods & Addons Copyright © 2017 DragonByte Technologies Ltd.
Resources saved on this page: MySQL 18.75%
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2017 DragonByte Technologies Ltd.
Copyright ©1999-2017 diyAudio