Wonderful picture, beatiful plate characteristics. Yet... I guess the Vg1= 0 line is the fifth from bottom, right? meaning the span in from Vg1 = -2,0 to Vg1 = +5,0 - if I'm correct. Now the question is... any value for G1 current for the positive region? how hard is this super-triode to drive?
and... mad question: did you ever try to trace these beam triodes with beam plates strapped to anode, maybe through a suitable resistor?
Yes, the Vg1=0 is the 5th line up. -2.0 to +5.0 Vg1 for that curve trace,
The 576 has an adjustable grid current limiter, 20 mA gets you up to the curve with a knee at 40 mA, and 100 mA gets you a curve with the knee at 200 mA.
I also checked using an external supply for grid1, and got the same results. You can see the limiting just kick in, in the knee, and then drop out as the plate V goes up. But the next curve up is held to the same, so looks like the grid current stays approximately at the knee value. So 50% of (knee value) plate current is absorbed by grid 1 thru the curve approx. +6.5Vg1 at 100 mA thru the grid (200 mA plate current) would be 0.65 Watt. Seems to tolerate that on the curve tracer. But the designers might not be happy with 0.65 Watt on their frame grid. The grid covers a fair bit of area over the large cathode, so the heat density may still be small compared to the cathode heat radiation ( 15.75 Watt heater ). One would want to do a Tubelab inspired test, up to destruction, to find a safe margin for amplifier use.
George! Got one of these tubes?
I tried connecting the beam plate to the plate, and get about 9 mA more plate current for the top curve. The grid1 current doesn't increase though, so that's good.
The knee voltages drop by about 20 V, helpful.
The 576 has an adjustable grid current limiter, 20 mA gets you up to the curve with a knee at 40 mA, and 100 mA gets you a curve with the knee at 200 mA.
I also checked using an external supply for grid1, and got the same results. You can see the limiting just kick in, in the knee, and then drop out as the plate V goes up. But the next curve up is held to the same, so looks like the grid current stays approximately at the knee value. So 50% of (knee value) plate current is absorbed by grid 1 thru the curve approx. +6.5Vg1 at 100 mA thru the grid (200 mA plate current) would be 0.65 Watt. Seems to tolerate that on the curve tracer. But the designers might not be happy with 0.65 Watt on their frame grid. The grid covers a fair bit of area over the large cathode, so the heat density may still be small compared to the cathode heat radiation ( 15.75 Watt heater ). One would want to do a Tubelab inspired test, up to destruction, to find a safe margin for amplifier use.
George! Got one of these tubes?
I tried connecting the beam plate to the plate, and get about 9 mA more plate current for the top curve. The grid1 current doesn't increase though, so that's good.
The knee voltages drop by about 20 V, helpful.
Last edited:
You mean digitizing the deflection data from the curve tracer?
Like this:
https://www.ebay.com/itm/303606230408
Only $5000 and no mods for doing tubes. They don't seem to offer the interface alone. I inquired as to the resolution, but they didn't respond. From some of the pics, it may only be 8 bit. The 576 unit looks in bad shape too.
I suspect one of those $70 micro computer boards with D/A A/D and USB interface could be programmed to do screen captures. Some additional work to do a fancy display and post processing of the data.
Last edited:
The Tek 576 display is usable as an X-Y display, that's all it does anyway. There are some internal relays that can switch the X-Y inputs to some external BNC connectors. ( I added the BNCs and control switches. ) Normally the High Current (200 Amp ) plug-in adapter used them. I have used the Vertical input to connect a DC/AC current probe for monitoring screen current during curve tracing.
To frame grab the display during curve tracing, one would just need two A/D channels on the uComputer board to capture simultaneous Vert. and Horiz. positions repeatedly, at a modestly high rep-rate (a few hundred points per plate curve) for however many plate curves are being done. Send those (cached in uP) over the USB link to an Extended Basic program on the PC to display the point collection on an X-Y graph. Could record data to a file, process parameters like mu, gm or rp at a cursor, plot load lines, Watt limits .... The uP board could also monitor user set safety limits for plate, screen or grid1 dissipations and use the D/A channel to limit the curve tracer excursion. Certainly do-able. Some programming work to do it professionally. Turning the 576 into a Tek 370/371 curve tracer essentially.
-----------------------------------------------
For those Beam Triodes, here are some pics of the plate with and without the "boxlets" (also called plate cavities ). You want the -without- "boxlet" case for lower plate voltage operation. These "boxlets" also occur on many of the later TV Sweep tubes. but there they had an extra fin or two inside, the "Barkhausen" fin or plate. That was used to stop VHF/UHF oscillations of charge around the screen grid. The "boxlet" and fin prevented much of the secondary electron emission by causing glancing electron strikes into a pocket or cavity (multiple bounces till absorption at low energy). The Beam Triodes were just Beam Sweeps with the Screen left out, so no issue with secondaries collecting around the Screen grid. But the "box", without the B fin, puts the plate further away, requiring higher plate V. Which was fine for TV HV limiting, but not good for DIY amplifiers.
Boxlets:
No "Boxlet": preferred
To frame grab the display during curve tracing, one would just need two A/D channels on the uComputer board to capture simultaneous Vert. and Horiz. positions repeatedly, at a modestly high rep-rate (a few hundred points per plate curve) for however many plate curves are being done. Send those (cached in uP) over the USB link to an Extended Basic program on the PC to display the point collection on an X-Y graph. Could record data to a file, process parameters like mu, gm or rp at a cursor, plot load lines, Watt limits .... The uP board could also monitor user set safety limits for plate, screen or grid1 dissipations and use the D/A channel to limit the curve tracer excursion. Certainly do-able. Some programming work to do it professionally. Turning the 576 into a Tek 370/371 curve tracer essentially.
-----------------------------------------------
For those Beam Triodes, here are some pics of the plate with and without the "boxlets" (also called plate cavities ). You want the -without- "boxlet" case for lower plate voltage operation. These "boxlets" also occur on many of the later TV Sweep tubes. but there they had an extra fin or two inside, the "Barkhausen" fin or plate. That was used to stop VHF/UHF oscillations of charge around the screen grid. The "boxlet" and fin prevented much of the secondary electron emission by causing glancing electron strikes into a pocket or cavity (multiple bounces till absorption at low energy). The Beam Triodes were just Beam Sweeps with the Screen left out, so no issue with secondaries collecting around the Screen grid. But the "box", without the B fin, puts the plate further away, requiring higher plate V. Which was fine for TV HV limiting, but not good for DIY amplifiers.
Boxlets:
No "Boxlet": preferred
Last edited:
The 370B curve tracer had A/D and D/A, measurement facility readouts, marker dots, cursors, etc., and GPIB (With respect to many of you, HPIB).
I think the last price from Tektronix (The 370B Sony Tektronix T&M Instrument), was US $38,000. Even at that price, production was finally shut down.
Sad Day.
I think the last price from Tektronix (The 370B Sony Tektronix T&M Instrument), was US $38,000. Even at that price, production was finally shut down.
Sad Day.
Well, if the used Tek 370... curve tracers on Ebay would come down out of the Stratosphere on pricing, I might consider trying one. As it is, however, the Tek 576 curve tracers on Ebay are getting very $$$ too. Seeing more of them cannibalized for parts now. Spare 576 parts (often worn out! ) are becoming more plentiful at least. Five years ago it was cheaper to buy a whole 576 for spare parts. I'm glad I got a good spare when it was cheap to do so.
That new Grid Step Amplifier module makes it easy to convert Tek 576, 577, 370, 371 Curve Tracers to tube use.
The new uP PC board curve tracers are leaving all this behind though.
--------------------------------------
Here is another rarely used TV Sweep tube, the 38/6HE7, usually quite cheap. This has a Beam pentode/tetrode.... and a damper diode in it. The 38HE7 often (about 80% of tubes around ) allows using the pentode without the damper powered up. (pins 10 and 12 for 21V pentode Htr) That drops enough heat out of the bottle to up-rate the B pentode to 15 Watt Pdiss (TV use rating, 20 Watt audio rating). The tube has nice pentode curves and nice triode curves.
That new Grid Step Amplifier module makes it easy to convert Tek 576, 577, 370, 371 Curve Tracers to tube use.
The new uP PC board curve tracers are leaving all this behind though.
--------------------------------------
Here is another rarely used TV Sweep tube, the 38/6HE7, usually quite cheap. This has a Beam pentode/tetrode.... and a damper diode in it. The 38HE7 often (about 80% of tubes around ) allows using the pentode without the damper powered up. (pins 10 and 12 for 21V pentode Htr) That drops enough heat out of the bottle to up-rate the B pentode to 15 Watt Pdiss (TV use rating, 20 Watt audio rating). The tube has nice pentode curves and nice triode curves.
Last edited:
The strangest thing on the Tek 576 was the discrete part D/A used for grid stepping. Originally it looked like I would just have to pull one IC pin to get the full SN7493 counter trace count for 15 traces. Which almost worked, except the D/A was configured for 1,2,4,4 instead of the more typical 1,2,4,8. Fortunately, it just took a resister change and a trimmer pot to get that to 1,2,4,8. What were they thinking? Some reason to prevent more traces? 15 traces still photographs well with my camera.
The other thing I didn't like was the 1500V scale. I don't like my hobby killing me. That got changed to a 750V scale (and 2x current) with a simple rectifier circuit mod. Before that, the toggle switch on the front doorstep for switching between tubes, clearly doesn't seem rated for 1500 V. It arced over after I tried 1500V, so that had to be fixed. Maybe one was supposed to disable collector voltage before toggling that switch. But I wanted fast toggling for comparing tubes.
Now if it only weighed less, so I could move it around. But with two working tracers now, I can put one on each bench. And when I found the 2nd 576 was in excellent condition, I ordered a 3rd cheap late model one for parts duty instead, which still needs some fixing up, but looks promising. Not sure if the floor can support 3 of them along with a bunch of scopes though. I might end up with the 1st tube amplifier that does a power-up curve trace tube test.
How much does the Tek 370B weigh?
The other thing I didn't like was the 1500V scale. I don't like my hobby killing me. That got changed to a 750V scale (and 2x current) with a simple rectifier circuit mod. Before that, the toggle switch on the front doorstep for switching between tubes, clearly doesn't seem rated for 1500 V. It arced over after I tried 1500V, so that had to be fixed. Maybe one was supposed to disable collector voltage before toggling that switch. But I wanted fast toggling for comparing tubes.
Now if it only weighed less, so I could move it around. But with two working tracers now, I can put one on each bench. And when I found the 2nd 576 was in excellent condition, I ordered a 3rd cheap late model one for parts duty instead, which still needs some fixing up, but looks promising. Not sure if the floor can support 3 of them along with a bunch of scopes though. I might end up with the 1st tube amplifier that does a power-up curve trace tube test.
How much does the Tek 370B weigh?
Last edited:
Tektronix Lawyers made sure that late designs of curve tracers had Caution warnings on the curve tracers, and in the curve tracer's manuals.
There were covers over the test connections and adapters, and the covers were interlocked, no power with the covers open or removed.
Only the adventuresome, stupid, and/or creative would bypass all this, and put their spouses at risk of being widows or widowers.
There were covers over the test connections and adapters, and the covers were interlocked, no power with the covers open or removed.
Only the adventuresome, stupid, and/or creative would bypass all this, and put their spouses at risk of being widows or widowers.
Almost all of the 576 curve tracers I've seen on Ebay have been missing that plastic safety box, and you can almost never find that box listed (went in the trash probably). The 2nd 576 I got (for parts originally), had pencil marks all around the safety switch activator location. It was clear that someone had been routinely using a pencil to override the safety feature. Since the unit would only erratically power up, I checked the relay activated by that switch, and the contacts were burned off from chattering. A real easy fix at least. Tek would be horrified at all the clip lead like connectors I use for connecting up tubes now. (which wouldn't fit in the safety box anyway. A toggle switch over-ride on the front DUT box now ).
I see the Tek 370B weighs 82 Lbs, more than the Tek 576 at 75 LBs. I can't imagine hauling one of those from a car to a customers lab.
https://w140.com/tekwiki/wiki/370
https://w140.com/tekwiki/wiki/576
I see the Tek 370B weighs 82 Lbs, more than the Tek 576 at 75 LBs. I can't imagine hauling one of those from a car to a customers lab.
https://w140.com/tekwiki/wiki/370
https://w140.com/tekwiki/wiki/576
Last edited:
The 370B covers, and higher power 371 with remote sockets and cover, had problems.
The problem that is often forgotten is the small space beneath the covers.
You could not fit a heat sink large enogh under the cover.
Even with low duty cycle pulsed operation, the high power solid state devices would explode without proper heat sinks.
And perhaps some of the covers were melted or blown up that way too.
Marketing could sell more curve tracers if the only photos in the Brochure had the covers off . . . more sexy looking.
But the Lawyers would never approve that.
The problem that is often forgotten is the small space beneath the covers.
You could not fit a heat sink large enogh under the cover.
Even with low duty cycle pulsed operation, the high power solid state devices would explode without proper heat sinks.
And perhaps some of the covers were melted or blown up that way too.
Marketing could sell more curve tracers if the only photos in the Brochure had the covers off . . . more sexy looking.
But the Lawyers would never approve that.
Ahh, I see TWO straps on top for carrying the 370B.
The 576 only has One!
Yeah, using heatsinks with the cover would not fit the 576.
For SS TO-220 or TO-247 .... I have the heatsinks sticking up in the air.
I should add the 4x 11 Lb power supplies for screen grids and Heaters to the 75 Lb weight of the 576 alone.
The 576 only has One!
Yeah, using heatsinks with the cover would not fit the 576.
For SS TO-220 or TO-247 .... I have the heatsinks sticking up in the air.
I should add the 4x 11 Lb power supplies for screen grids and Heaters to the 75 Lb weight of the 576 alone.
Last edited:
About a year ago, maybe a little less, the Tek museum was looking for homes for two working, refurbed 576's for $100 each, but must pick up at the museum. Now I'm feeling guilty about not splashing it onto diyAudio. And sorry I didn't get my friend in PDX to pick one up for me. Oh well.
All good fortune,
Chris
All good fortune,
Chris
When I think back (like almost 20 years now) to when I got the original Tek 576 here, I'm still a little puzzled by the event. I think I was looking at tube testers on Ebay and was shocked that the fancier ones were up around $500. Then there was this Tek 576 listed for $600 in "good working condition", which seemed like a no brainer at the time. The thing could do 1500V after all.
After I got it, in a shipping crate, I began to realize the limitation was the grid stepping voltage range, which was only 20V. So that got me working on a boosted step amplifier, which got boosted more and more along the way ( so I could do +grid 2 curves as well ). Then pentode screen grids were a problem, so that got the wiring mod to the selector toggle switch, with an external power supply. Then some front panel push-buttons were very flaky, so I figured out how to re-new them like new. And finally the power supply electrolytics started getting weak, so on and on. At least it was reasonably calibrated. It was operating like a new unit by then with impressive capabilities for tubes.
Later, after watching some YouTube videos on the 576, I began to appreciate some of the finer points on repairs, and what to look for when buying one. Turned out I have one of the earliest ones made from the 1st year of production in 1968. But the HV power supply and the CRT had been upgraded to the latest revisions. I guess the originals burned out. As an early unit, it has sockets for the ICs and semiconductors with gold plated circuit boards and high quality trimpots. The later models don't have the IC sockets, no gold plating and cheezo trimpots that fall apart. After I got a late model unit (for parts originally) I found all this cost cutting stuff. At least it had good electrolytics and good push-buttons. If the oldie unit didn't have the IC sockets, I wouldn't have tried the mod to get 15 traces.
Anyway, both are working smoothly now. Well calibrated. Cosmetically good.... Fun to operate.
A powerful tool for trying new variations, like Crazy Drive, or grid 2 drive, Schade Fdbk, UnSet, ..... (lots more)
When George mentioned he had a new secret scheme that he had been working on for a year (UnSet etc) , which gave better Fdbk and power, it only took me 5 minutes to come up with the same thing. Just flipped the grid step drive to the cathode and the cathode ground connection to the grid1. Then added the Schade R divider to Gnd, for grid 1. Same results. Turned out we both had the same thing.
So many unusual tubes out there to trace.
9KC6 grid 3 drive curves:
Crazy Drive on a 26LX6:
The Perfect Triode:
The Perfect Pentode:
6HJ5 with UnSet:
After I got it, in a shipping crate, I began to realize the limitation was the grid stepping voltage range, which was only 20V. So that got me working on a boosted step amplifier, which got boosted more and more along the way ( so I could do +grid 2 curves as well ). Then pentode screen grids were a problem, so that got the wiring mod to the selector toggle switch, with an external power supply. Then some front panel push-buttons were very flaky, so I figured out how to re-new them like new. And finally the power supply electrolytics started getting weak, so on and on. At least it was reasonably calibrated. It was operating like a new unit by then with impressive capabilities for tubes.
Later, after watching some YouTube videos on the 576, I began to appreciate some of the finer points on repairs, and what to look for when buying one. Turned out I have one of the earliest ones made from the 1st year of production in 1968. But the HV power supply and the CRT had been upgraded to the latest revisions. I guess the originals burned out. As an early unit, it has sockets for the ICs and semiconductors with gold plated circuit boards and high quality trimpots. The later models don't have the IC sockets, no gold plating and cheezo trimpots that fall apart. After I got a late model unit (for parts originally) I found all this cost cutting stuff. At least it had good electrolytics and good push-buttons. If the oldie unit didn't have the IC sockets, I wouldn't have tried the mod to get 15 traces.
Anyway, both are working smoothly now. Well calibrated. Cosmetically good.... Fun to operate.
A powerful tool for trying new variations, like Crazy Drive, or grid 2 drive, Schade Fdbk, UnSet, ..... (lots more)
When George mentioned he had a new secret scheme that he had been working on for a year (UnSet etc) , which gave better Fdbk and power, it only took me 5 minutes to come up with the same thing. Just flipped the grid step drive to the cathode and the cathode ground connection to the grid1. Then added the Schade R divider to Gnd, for grid 1. Same results. Turned out we both had the same thing.
So many unusual tubes out there to trace.
9KC6 grid 3 drive curves:
Crazy Drive on a 26LX6:
The Perfect Triode:
The Perfect Pentode:
6HJ5 with UnSet:
Last edited:
Here you go, a 175 Lb Curve Tracer, the Iwatsu CT-3200 with CS-302 Test Fixture on Ebay. 3000V max, 1000A max, Ethernet and USB interfaces, Graphical Color Display. Apparently New! $ 15000 plus shipping. BUT, only +/- 40 V stepping range. (according to their Website, but I do see a Base/Gate terminal marked 300V pk? )
Only 40V stepping! "No Operator Serviceable Components Inside" Darn. I'll have to pass on this. A little over $400 budget too.
But look at how big their safety test box is!
Could fit tubes in there. But does it have a double test port for comparing/matching devices?
I'll bet it has a cooling fan too.
https://www.ebay.com/itm/276428575300
Only 40V stepping! "No Operator Serviceable Components Inside" Darn. I'll have to pass on this. A little over $400 budget too.
But look at how big their safety test box is!
Could fit tubes in there. But does it have a double test port for comparing/matching devices?
I'll bet it has a cooling fan too.
https://www.ebay.com/itm/276428575300
Last edited: