uTracer (Micro Curve Tracer)

General discussion area for tube amps.

Moderators: pompeiisneaks, Colossal

Post Reply
User avatar
martin manning
Posts: 13080
Joined: Sun Jul 06, 2008 12:43 am
Location: 39°06' N 84°30' W

Re: uTracer (Micro Curve Tracer)

Post by martin manning »

jazbo8 wrote:I was wondering if you have a 7199 handy? I am looking for the triode-strapped plate curves for the pentode section, it seems that no one has done one yet...
No, sorry, I don't have one.
User avatar
jazbo8
Posts: 311
Joined: Sat Jul 23, 2011 12:59 am

Re: uTracer (Micro Curve Tracer)

Post by jazbo8 »

martin manning wrote:No, sorry, I don't have one.
No worries, it is not a popular tube anymore and getting harder to find...
User avatar
martin manning
Posts: 13080
Joined: Sun Jul 06, 2008 12:43 am
Location: 39°06' N 84°30' W

Re: uTracer (Micro Curve Tracer)

Post by martin manning »

Playing with the uTracer again today...

The first plot is an EH 6V6GT traced at 400V plate and screen, which is the approximate operating point in a Fender Deluxe Reverb. Plate curves at 400V Vg2 will not be found in vintage 6V6 spec sheets, but this particular tube measures very close to the spec at 250V (112% Ia, 88% Is, 103% gm, and 100% Ra), so it is a very good specimen. An 8k a-a load line at 400V Va (as used in the DR) will be a bit below the knee, but likely right through it with a 100% spec tube and including the screen resistor effect.

This 6V6 trace is also right at the limit of the uTracer's capability for a full trace at 400V Vg2. The highest point on the Vg1=0 curve at ~400V and ~230 mA requires turning the software limit off (the nominal max current is 200 mA). This is a relatively low risk since I know that the tube is not shorted, and there is still a hardware limit.

The second plot shows the effect of several screen resistor values on a 6L6GC. I did something similar earlier in this thread, but this shows the effect a little better and for a range of resistance (R = 0, 470, 1k, 1k5, and 2k). With a 2k Rg2, the max grid current is reduced by 50% and the Ia at the knee is pulled down about 15%, which has some implications for the choice of load impedance.
You do not have the required permissions to view the files attached to this post.
User avatar
martin manning
Posts: 13080
Joined: Sun Jul 06, 2008 12:43 am
Location: 39°06' N 84°30' W

Re: uTracer (Micro Curve Tracer)

Post by martin manning »

Hardware mods again.

From the beginning I have been using an FTDI 5V USB-TTL adapter cable, which requires removing the uTracer's RS232 interface IC and jumpering over its socket. I selected this cable for its compact 3.5mm TRS audio plug termination, and it is a high-quality part. With this set up I noticed that the "Power On" LED was dimly lit by the USB connection with the uTracer power off, and the "HV On" LED would be dimly lit if the power was shut off after running a test. I haven't had any problem using it this way, but R Dekker notes that this might result in excess current draw and damage to the adapter cable. This is good to know, and in my case the recommended fix is easy: The Rx (receive) jumper is simply replaced with a 1k8 resistor.

Ronald also notes a potential stability issue with leaving the controller's Rx terminal floating, which would occur if the cable is disconnected with power on. He recommends an 18k pull-up resistor from the Rx terminal to +5V, which I soldered to the bottom of the PCB under the DIP socket, pin 9 to pin 16.

Everything seems to be functioning properly- the Power On LED does not glow when it shouldn't, and I only see a brief flash on the HV LED as the uTracer power is switched off. These modifications preserve the RS232 interface socket and the original 3-pin polarized header connection, so all options are still open with a minimal effort.
You do not have the required permissions to view the files attached to this post.
User avatar
martin manning
Posts: 13080
Joined: Sun Jul 06, 2008 12:43 am
Location: 39°06' N 84°30' W

Re: uTracer (Micro Curve Tracer)

Post by martin manning »

6SN7GTB Cascode (off the beaten path for guitar amps...)

As seen in the triode-connected trace at the upper-left, this NOS Tung-Sol is a remarkable tube- the two sections are virtually identical, and spot-on the spec for anode current, gm, ra, and mu!

The upper-right plot is a trace for the cascode connection, which is measured as a pentode with the upper grid being treated as the screen. A lot of grid current flows in the upper triode at anode voltages below the upper grid voltage, and two "knees" can be seen in the Vg1=0 curve: one where the screen current begins to level-off (much like in a pentode), and another at 2x the upper grid voltage where the screen current cuts off. In tubes where the grid current is very limited on the negative side of Vg-k=0, the upper knee consistently appears at anode voltage equal to 2x the upper grid voltage. The two triodes have the same current, and at that point each one has Va-k equal to the upper grid voltage since Vg-k=0.

The upper grid voltage is usually set by a high-resistance voltage divider at about 1/4-1/2 of the HT, which will severely limit the upper grid current. The two lower plots were made with a 10k and a 100k resistor in series with the upper grid to simulate that effect progressively. Note that the upper knee on the Vg1=0 line remains at a fixed current and at upper anode voltage equal to 2x the upper grid voltage. The current there is as found on the triode Vg1=0 curve at Va-k=upper grid voltage, while the lower knee is straightened out. This makes it easy to predict the location of the upper knee for choosing a load line, and the cutoff grid voltage and bias point can be found from the triode curves at the chosen upper grid voltage.
You do not have the required permissions to view the files attached to this post.
User avatar
rogb
Posts: 1078
Joined: Fri Apr 02, 2010 9:56 am
Location: London, England

Re: uTracer (Micro Curve Tracer)

Post by rogb »

My uTracer is finally up and running!
Here are a few photos of the uTracer in action. The case is constructed according to a superb Martin Manning design and is a Hammond 1590D so very compact!

After a couple of problems that were entirely my own fault, forgetting to limit current, so I had to send the board back to Ronald to replace the SMDs, and switching off the compliance(!), here you can see the uTracer in action measuring a 12AX7 dual triode. 12A*7s have their own hardwired socket as I test a lot of these. The other tube is a Brimar 6V6GT which tests perfectly.

I also have tested a pair of 6V6GTAs supplied to me for a Fender 5E3 build and they match almost exactly!

I have used the internal heater supply as I detected no difference to using an external supply with the sort of common tubes I test. Maybe it might be an idea to use an external DC supply for EL34s though.
This is a great piece of equipment, it will improve your fine pitch soldering no end and even for me was quite easy to construct following the excellent manual and online help, and not a few emails between Ronald, Martin and myself.
You do not have the required permissions to view the files attached to this post.
User avatar
martin manning
Posts: 13080
Joined: Sun Jul 06, 2008 12:43 am
Location: 39°06' N 84°30' W

Re: uTracer (Micro Curve Tracer)

Post by martin manning »

Looking good Rog, nice to see another one come to life here!

Clever idea to have a dedicated 12A_7 socket... those require the most complex test setup using the patch cords so you'll definitely save yourself some time there. I recommend adding a continuous mode switch. It's very handy for checking calibration and testing the system without the need to open the enclosure.

The internal heater supply is nominally rated to 1.5A, so large power tubes (EL34's, 6CA7, 6550/KT88, etc.) are better powered using an external supply.
pdf64
Posts: 2688
Joined: Sat Mar 12, 2011 9:23 pm
Location: Staffordshire, UK

Re: uTracer (Micro Curve Tracer)

Post by pdf64 »

Wow, nice work, well done!
User avatar
martin manning
Posts: 13080
Joined: Sun Jul 06, 2008 12:43 am
Location: 39°06' N 84°30' W

Re: uTracer (Micro Curve Tracer)

Post by martin manning »

Here's a series of traces of the Vg1=0 curve for a range of filament voltages on a 6V6. The top three curves are +10%, nominal, and -10%, and it's clear that there is a relatively weak effect on anode current within that range. The cliff doesn't occur until the voltage drops to more than 30% below nominal, ~4.4V.
You do not have the required permissions to view the files attached to this post.
User avatar
rogb
Posts: 1078
Joined: Fri Apr 02, 2010 9:56 am
Location: London, England

Re: uTracer (Micro Curve Tracer)

Post by rogb »

Thanks Martin, that's interesting.
I've got some new JJ6V6 here and I'll do some Vh sweeps to see how they hold up.
User avatar
rogb
Posts: 1078
Joined: Fri Apr 02, 2010 9:56 am
Location: London, England

Re: uTracer (Micro Curve Tracer)

Post by rogb »

The JJ 6V6s stayed good till 5.3v then fell off the cliff, not as good a result as yours.
I'll try the NOS 6V6 soon. Interesting to see how tolerant of heater voltage variation your tubes are.
Also today got an email from Ronald for a beta update for the software.
I ran it the executable but it wiped my cal file and I couldn't get it to test at all. Luckily I had a backup of my cal file so all was well.
User avatar
martin manning
Posts: 13080
Joined: Sun Jul 06, 2008 12:43 am
Location: 39°06' N 84°30' W

Re: uTracer (Micro Curve Tracer)

Post by martin manning »

The tube in the heater voltage sweep above is a recent EH. Here's a JJ, which does show a little less tolerance for low Vf.

I suspect that different cathode coating materials and perhaps different filament alloys are responsible, and I noticed the filament current is higher for the JJ (0.54A vs. 0.47). I suppose as long as the performance at +/- 10% from nominal is good, and it is, then they meet the expected spec. The JJ's plate area is probably 2x the EH's- no surprise they will take more dissipation!

I have the new GUI release up and running too; no issue with the cal file here. Some of those annoying things are just due to different versions of the Windows OS. For example I never experienced the bug Ronald mentioned (failure to plot saved curves when opening a setup that has them in it). I'm running Windows 8 (on my MacBook), BTW. For this kind of update you only need to drop the executable into the folder that contains the previous version(s). It will find and use the existing cal file.
You do not have the required permissions to view the files attached to this post.
User avatar
rogb
Posts: 1078
Joined: Fri Apr 02, 2010 9:56 am
Location: London, England

Re: uTracer (Micro Curve Tracer)

Post by rogb »

The GUI update now runs fine on Windows 8.1 if you right click on the .exe and Run As Administrator, copying the cal file perfectly :)
User avatar
rogb
Posts: 1078
Joined: Fri Apr 02, 2010 9:56 am
Location: London, England

Re: uTracer (Micro Curve Tracer)

Post by rogb »

Here are some plots, some Vh curves showing the range of tolerance of some tubes and also a plot from Nick's new GUI alternative to Ronald's, uTmax, this is in alpha testing so there may be some glitches with the software and calibration, also the release notes need to be updated, but it works on the 400v uTracer 3+ and is very elegant :)
http://www.bmamps.com/Tech_uTgui.html
You do not have the required permissions to view the files attached to this post.
potatofarmer
Posts: 50
Joined: Fri Dec 02, 2011 2:54 am

Re: uTracer (Micro Curve Tracer)

Post by potatofarmer »

It's not a 7199 but I have a ton of 6GH8s lying around. I could donate a couple if martin is willing/interested in checking out a triode/pentode preamp tube.
Post Reply