Pages: 1 2 3 4 5 6 7 8 9 10 11 12 13
Nick you could adapt it for pentodes by adding another HT pot to supply the screen and another voltmeter to set the screen voltage, if the suppressor has a separate pin connection then that would be connected to the cathode, plot each Ia curve referenced to a fixed screen voltage.
Lawrence.
Ah. I think I understand what you're saying Lawrence.
Over the course of the next couple of days I'll have a go at a revised circuit and I'll then post it for consideration
Thanks.
Nick.
I've always had regulated power supplies around, for HT, and I've always used similar supplies for grid bias. So the tracer itself contains only sockets, switching, and perhaps meters. I've built a couple. I used X-Y plotters for the permanent record; they're also faster than reading meters by eye, so you can run tubes in overload for very short times and get points on the curve you couldn't normally. These days you can't give away X-Y plotters and I have several. I also played around with CRT tracers. But that was all years ago when I was writing my book on test gear.
Hi Nick, that sounds a good project. There are things like Lab Jacks that interface (isolated) to a PC and are low cost. They dump data into an XL file and graphs can be plotted directly from that.
It should also be possible to control the Sussex PSUs from a PC to step voltages etc.
I have a curve plotter circuit here from Electronic Engineering in the late 40's. It uses X-Y plotters and uniselectors to step the voltages.
It should make a good winter project if yoyu are not too busy with the house.
Plenty of ideas for projects in my tech library when you get over.
Ed
I've had a uTracer for about 20 months now. Wonderful bit of kit.
Not only does it handle pretty much every tube type with ease, it also has really neat tricks like plotting both halves of a double triode simultaneously for matching, testing magic eyes and generating and exporting SPICE models for the DUT.
Brilliant!
[
attachment=12434]
(15-07-2015, 07:38 PM)Ed_Dinning Wrote: [ -> ]Hi Nick, that sounds a good project. There are things like Lab Jacks that interface (isolated) to a PC and are low cost. They dump data into an XL file and graphs can be plotted directly from that.
It should also be possible to control the Sussex PSUs from a PC to step voltages etc.
If you are going to use a PC to control the voltages of a standard old-school valve tester, then use USB-connected data acquisition and export in to Excel, why not make your life a WHOLE lot simpler, cut out all the middle-man stuff and use a
uTracer 3+
The uTracer it almost entirely through-hole (the small surface mount stuff is done for you before you get it) and takes just a few hours to build, all a single Eurocard format (10cm x 16cm) - the kit is very high quality with only top quality components used. The price is set to only cover costs, not to make a profit. It's a labour of love... an active project under continuous development...
The designer,
Ronald Dekker, is a professor at Tu Delft & Philips - he has fully documented (in minute detail) every design decision, including failed approaches, complete with detailed explanations. He has a long-standing interest in valves - his website is a goldmine of historical information as he has access to all of Philips' archives.
[
attachment=12435]
It'll almost certainly cost less, is far far simpler, and way more powerful than any lash-up that you can put together yourself. You'll also learn a ton of stuff building and using it. I can bung a valve into it and have a set of graphs of ACTUAL, not theoretical, behaviour at my chosen operating points in just a few minutes (typically under 5) - over 400 have been built, world-wide
Just some of what it does (there is a whole bunch more, including automatic THD calculation for chosen operating points etc.) - also see
What is the uTracer and What can it do?":
- Overview:
- pulsed measurement of diodes, triodes, pentodes, etc.
- 13 pre-defined measurements:
- Ia,Is versus Vgrid; stepping Vanode; Vscreen & Vheater constant
- Ia,Is versus Vgrid; stepping Vanode=Vscreen; Vheater constant
- Ia,Is versus Vanode; stepping Vgrid; Vscreen & Vheater constant
- Ia,Is versus Vanode; stepping Vscreen; Vgrid & Vheater constant
- Ia,Is versus Vanode=Vscreen; stepping Vgrid; Vheater constant
- Ia,Is versus Vscreen; stepping Vgrid; Vanode & Vheater constant
- Ia,Is versus Vscreen; stepping Vanode; Vgrid & Vheater constant (positive grid bias mode)
- Ia,Is versus Vanode; stepping Vscreen; Vgrid & Vheater constant (positive grid bias mode)
- Ia,Is versus Vheater; stepping Vgrid; Vanode & Vscreen constant
- Ia,Is versus Vheater; stepping Vanode; Vgrid & Vscreen constant
- Ia,Is versus Vanode, stepping Vg with Vs = UL(Va,k) simulating ultra-linear operation for pentodes
- Ia,Is versus Vgrid, stepping Vg with Vs = UL(Va,k) simulating ultra-linear operation for pentodes
- Ia,Is versus Vanode; stepping Vgrid; Vscreen & Vheater constant; simulating “Schade” feedback
- plot transconductance (Mutual Conductance) as function of bias.
- plot output resistance as function of bias
- continuous measurement of magic eyes
- matching of double-triodes (second anode connected to screen terminal)
- Quick Test feature which measures: Inom, Rp, Gm and mu in a matter of seconds without the need to plot full curves!
- ...
- Anode and Screen voltage sources.
- 2 to 400 V, resolution 0.4 V
- measurement pulse 1 ms.
- nominal output current 200 mA.
- maximum output power 60 W
- hardware current limit 250 mA.
- programmable compliance: 25, 50, 75, 100, 125, 150, 175, 200 mA, response time 20 us.
- short circuit proof.
- linear / logarithmic sweep
- Anode and Screen current measurement
- 0 to 200 mA in 8 ranges with 9.5 bit resolution
- fully auto ranging or manual range selection
- auto averaging or manual average selection
- ranges: 0-1 mA, 0-2 mA, 0-5 mA, 0-10 mA, 0-20 mA, 0-40 mA, 0-100 mA, 0-200 mA.
- Grid bias; negative
- 0 to -50 V, resolution 50 mV
- protected against short circuits
- linear / logarithmic sweep
- Grid bias; positive
- triodes only (screen supply is used source)
- 0 to 200 V
- 0 to 200 mA
- pulsed, so no grid dissipation
- Heater
- 0 to Vsupl (nom. 19.5 V).
- higher voltages with external supply.
- 1.5 A max, limited by fuse.
- direct and indirectly heated cathodes.
- soft start, 0-100% in 10 seconds
- external power supply can be used for higher voltages or low voltage / high current heaters
- Data representation
- On screen graphic presentation with cursor function
- Automatic or manual scaling of axis
- 2 to 25 order polynomial fit though data for extraction of transconductance or output resistance.
- Store graph to file in bmp format
- Store data in table format (several formats possible)
- Storage of all tube specific measurement settings!
- ...
I am sure the uTracer is a fabulous piece of kit, but at 215 euros it isn't cheap and at present I don't need all that functionality and I may never need it. My budget is more like £30 quid and this thread is about designing and building a simple [static] valve curve plotter for this sort of budget.
To clarify the objective, it is: "To study the characteristics of triode and pentode valves manufactured pre-1936 and determine their constants by the characteristics."
Specification:
1. Ability to measure and plot the plate characteristics and the transfer (mutual) characteristics.
2. Tester to handle triode and pentode valves produced before 1936.
3. Be able to accommodate Battery and AC valves.
4. Handle valves with 1.5v, 2v, 4v, 6v and 6.3v heaters, anode voltages up to 250v, and GB voltages down to -30v
5. Tester to be powered from the UK mains supply.
6. Simple switch to alternate between triode and pentode valves.
7. Selector switch to select appropriate heater voltage.
8. Adjustable anode and grid voltages.
9. Current readings from integral milliammeter and plotted manually in Microsoft Excel.
This is more complex than I originally envisioned, but I believe it is within my ability and my pocket and that the functionality is likely to suffice for the receivers I plan to restore and build.
I do have one question.
From what I have read I believe it's acceptable for the purpose of these tests to use a DC only supply, even for AC valves. Is this right?
It will certainly make the circuit simpler if it is.
Thanks
Nick
(16-07-2015, 11:25 AM)SurreyNick Wrote: [ -> ]I am sure the uTracer is a fabulous piece of kit, but at 215 euros it isn't cheap and at present I don't need all that functionality and I may never need it. My budget is more like £30 quid and this thread is about designing and building a simple [static] valve curve plotter for this sort of budget.
My comments were slightly tongue-in-cheek - like many, I do mainly audio with some RF on the side, so for most folk (like me), the uTracer is ideal. It is also likely to be pretty much the same cost as a Sussex or similar self-build manual "traditional" tester.
With a very specific requirements list, a targeted rather then general-purpose bit of hardware may well be better for you... however, if in the future your ambitions expand, what then
Quote:From what I have read I believe it's acceptable for the purpose of these tests to use a DC only supply, even for AC valves. Is this right?
DC is just fine...
Yes Nick, you can use DC for the heater supply for indirectly heated cathode valves.
The problem you might run into is the complexity of the unit for plotting the curves such as different valve bases and switching needed etc.
The usual trick to overcome some switching is to wire the valve sockets in parallel, pin 1 to pin 1 to pin one etc, pin 2 to pin 2 to pin 2 etc and bring those connections out in the form of fly leads with plugs on the end and numbered, these can then be plugged into the voltage supply sockets on the front panel marked as C, G1, G2, G3, A, Htrs etc, you would need to compile or refer to a set of data for the connections. It might be wise to decouple the valve socket pins to HT -ve as some valves can have the propensity to turn themselves into oscillators
Remember the unit as such is not a full tester, it could be adapted to test for leakage, shorts and gas etc it depends how far you want to go.
This post might cross with others, it took me a while to write it.
Lawrence.
Pages: 1 2 3 4 5 6 7 8 9 10 11 12 13