Keyboard Shortcuts
ctrl + shift + ? :
Show all keyboard shortcuts
ctrl + g :
Navigate to a group
ctrl + shift + f :
Find
ctrl + / :
Quick actions
esc to dismiss
Likes
Search
Infamous D180 is it good or bad.
|
I have a Tek 284 that I am working on what is mostly working there is just some issues with the pulser output of course. I finally worked up the nerve tonight to test the Diode it has been on the shelf for a while. I knew these didoes were fragile and are very expensive to replace. I found them¡ for $250.00 each diode. So the way I have things set up, a curve tracer set to DC and very sensitive measurements. I have the collector going out to a Keithly 6500 to watch the current. As I turn up the collector I get current flow in to the diode. At about 9mA there is a LARGE increase in current. But NOT a short. If I dial down the collector it looks like the diode recovers. RIGHT about 9 mA I can get it to flutter between the 2 current readings. So if I am reading the instruments correctly is this tunnel diode good? I am thinking it is but I am very new to tunnel diodes. I know I am in the deep end of the pool with this one but confirming the diode would be great news to getting this 284 up and running. I can put up a short video if it will make more scene
Eric |
|
Hi Eric,
toggle quoted message
Show quoted text
Download a short article on line about how tunnel diodes work and you will be able to tell from the pattern on the curve tracer how to confirm it is working and you can actually measure several important TD parameters. The negative resistance region of the TD is not something Tek curve tracers handle very well so you will probably see something like oscillation on the trace. That is OK. The TD will go to 4 different points on the CRT as the voltage and current changes around the area where the negative resistance is. The collector voltage setting should be on the lowest position of your curve tracer (6V on a 577). The horizontal volts/Div should be set at its lowest setting (50mV/Div on a 577). The vertical current should be set to 2mA/Div if you have a 10mA TD. The series resistor should be fairly low like 120 ohms on my 577. These settings will get you in the ball park. Very slowly increase the collector voltage. As you reach 50mV the collector current may have already reached 9mA. If it has you should start over. If it hasn't yet reached 9ma then continue increasing the voltage towards 100mV. At some point, the current which has been rising in almost a straight line will slow down then stop rising. That point marks the peak current or Ip. This is an important parameter to document. You should also make note of the voltage at the peak current. That is the peak voltage or Vp. If you attempt to increase the voltage beyond this point the tunnel diode enters its negative resistance region where it drives the curve tracer crazy. The tunnel diode will suddenly shift a few hundred mV to the right and the current will drop by up to half. It is now in a new stable location. Don't bother to increase the voltage any further. To find out everything else you need to know you have to start reducing the voltage very slowly. As you do this the current will begin to drop. At a certain point the current will slow down and stop dropping. This may be around 1mA. This is the valley current or Iv for your TD. The voltage at that point is Vv or valley voltage. If you reduce the voltage a bit more the TD will now jump again. This time it will be almost right back where you started. It may be 1mA up on the curve and you started at 0mA but that is not important. You are at the point where you can begin the entire cycle all over again. Now you have Ip, Vp, Iv, and Vv so you can calculate the approximate negative resistance of your TD. Rtd = (Vp - Vv) / (Ip - Iv). Dennis Tillman W7pF -----Original Message-----
From: [email protected] [mailto:[email protected]] On Behalf Of Eric Sent: Saturday, June 06, 2020 6:15 PM To: [email protected] Subject: [TekScopes] Infamous D180 is it good or bad. I have a Tek 284 that I am working on what is mostly working there is just some issues with the pulser output of course. I finally worked up the nerve tonight to test the Diode it has been on the shelf for a while. I knew these didoes were fragile and are very expensive to replace. I found them¡ for $250.00 each diode. So the way I have things set up, a curve tracer set to DC and very sensitive measurements. I have the collector going out to a Keithly 6500 to watch the current. As I turn up the collector I get current flow in to the diode. At about 9mA there is a LARGE increase in current. But NOT a short. If I dial down the collector it looks like the diode recovers. RIGHT about 9 mA I can get it to flutter between the 2 current readings. So if I am reading the instruments correctly is this tunnel diode good? I am thinking it is but I am very new to tunnel diodes. I know I am in the deep end of the pool with this one but confirming the diode would be great news to getting this 284 up and running. I can put up a short video if it will make more scene Eric -- Dennis Tillman W7pF TekScopes Moderator |
|
Hi Dennis,
toggle quoted message
Show quoted text
I checked the curve tracer this morning. Vc was 6.5V 8K series resister, vertical was 0.05 V/Div. Horizontal was 5uA/Div My confusion is I am not getting the typical curve that I have seen with tunnel diodes and a curve tracer. I took a picture this morning of the curve tracer. I had horizontal magnification turned on. This is right around 10mA as measured on the meter. I do now see the gap in the diode. Does the curve look somewhat strange because this one is just so fast. Also if the collector is increased just a little more the diode just turns ON and stops switching. I would test a few more to see if they all behaved the same however I have a sample size of exactly one. From the reading I have done and my lack of understand I am leaning towards this is a good diode and my problems with the 284 are bias related. Photo has been posted here /g/TekScopes/album?id=248216 Eric On 6/7/2020 3:17 AM, Dennis Tillman W7pF wrote:
Hi Eric, |
|
On Sun, Jun 7, 2020 at 02:50 PM, Eric wrote:
Just dropping in so ignore if I'm telling nonsense. 6.5V with a series resistor of 8k won't result in 10mA or even 1mA. Dennis gave some sensible values and *do* watch Alan Wolke's video at , as pointed to by another member a while ago. Raymond |
|
Hi Erik,
toggle quoted message
Show quoted text
Your description is quite confusing to me. 6.5V/8k is < 1mA, far below Ip (which is 21 mA ! for this type of TD). So I would expect to see no oscillation at all (unless perhaps leads pick up a lot of hum). What are exactly the V/div and mA/div in the picture itself (as result of 10X or whatever)? Did you really show the current horizontally? Anyway, since your TD oscillates it seems to be still functional. The usual pictures are taken in NPN mode (not DC mode) and using a much smaller series resistor. Here the series resistor should be small enough to deliver over 21 mA at highest Vc, but for safety not too small in the first attempts. I displayed vertically and V horizontally. There can/will be oscillation during state switches but the large picture will be clear. You will be able to read Ip from the picture. It is difficult and not necessary here to show the whole continuous curve including the negative resistance region. Did you dismantle the TD housing? It should be possible to test the TD in place via the output connector. Normally if you turn the internal Bias pot from one extreme to the other you should see the state switches of the TD at the output connector voltage. Normally the TD is stable in its Low state in one extreme and stable in its Hi state in the other extreme position. One possibility is that you can't obtain the Hi state because the bias current is not high enough. Or to say it better, because Ip of the TD has increased over the years. (Try also with the front panel Bias adj in the highest current position, I think that's clockwise.) Temperature also plays a role in Ip. IIRC I cheated with the Calibration in that I adjusted the regulated +/- 20 V supply slightly higher than specified. That increases the Bias current. Albert On Sun, Jun 7, 2020 at 02:50 PM, Eric wrote:
|
|
FYI, I uploaded an image of a 20mA Tunnel diode as shown on a Tek 576 curve tracer into your album
/g/TekScopes/album?id=248216&p=Created,,,20,2,0,0 Raymond |
|
Hi Raymond,
toggle quoted message
Show quoted text
Nice picture without oscillations. Which series resistance did you use in the 576? Did you connect the TD to the 576 with as short as possible leads? Albert On Sun, Jun 7, 2020 at 10:52 PM, Raymond Domp Frank wrote:
|
|
Hi Raymond,
toggle quoted message
Show quoted text
I was also impressed with your TD photo. That is the cleanest TD curve I have seen on a curve tracer. Dennis Tillman W7PF -----Original Message-----
From: [email protected] [mailto:[email protected]] On Behalf Of Albert Otten Sent: Sunday, June 07, 2020 2:27 PM To: [email protected] Subject: Re: [TekScopes] Infamous D180 is it good or bad. Hi Raymond, Nice picture without oscillations. Which series resistance did you use in the 576? Did you connect the TD to the 576 with as short as possible leads? Albert On Sun, Jun 7, 2020 at 10:52 PM, Raymond Domp Frank wrote:
-- Dennis Tillman W7pF TekScopes Moderator |
|
Hi Albert and Dennis,
I took the photograph some 5 years ago and don't remember doing anything special to get the picture. Maybe I put a cap across, lowest voltage range, no extra series resistance. Since I still have the same 576 and that particular TD should be around as well (or I'll take another), I'll try and repeat in the next few days and report back. Raymond |
|
I completely agree with Albert's comments.
toggle quoted message
Show quoted text
Your vertical was 5uA/Div and your horiz was 0.05V/Div. Not what you wrote. I suggested a vertical setting of 2ma/Div. Because the series resistor you chose was 66X larger than what I suggested you had to set the vertical 400x lower than I said suggested. A straight line (like the one in your photo) is exactly what you get from connecting a resistor across the curve tracer inputs. Since you have two of these straight lines it might indicate there is an oscillation back and forth between them Otherwise the photo you posted is useless. ** Take another photo and follow my directions or we won't be able to help you. The 576 has different series resistors than the 577 I have. I suggested 120 ohms but for the 576 choose the closest value that less than 200 ohms or greater than 50 ohms. Albert said you could start at a lower resistance but risks blowing out the TD since it seems like you don't like following directions. ** Dennis Tillman W7pF -----Original Message-----
From: [email protected] [mailto:[email protected]] On Behalf Of Eric Sent: Sunday, June 07, 2020 5:50 AM To: [email protected] Subject: Re: [TekScopes] Infamous D180 is it good or bad. Hi Dennis, I checked the curve tracer this morning. Vc was 6.5V 8K series resister, vertical was 0.05 V/Div. Horizontal was 5uA/Div My confusion is I am not getting the typical curve that I have seen with tunnel diodes and a curve tracer. I took a picture this morning of the curve tracer. I had horizontal magnification turned on. This is right around 10mA as measured on the meter. I do now see the gap in the diode. Does the curve look somewhat strange because this one is just so fast. Also if the collector is increased just a little more the diode just turns ON and stops switching. I would test a few more to see if they all behaved the same however I have a sample size of exactly one. From the reading I have done and my lack of understand I am leaning towards this is a good diode and my problems with the 284 are bias related. Photo has been posted here /g/TekScopes/album?id=248216 Eric On 6/7/2020 3:17 AM, Dennis Tillman W7pF wrote: Hi Eric, -- Dennis Tillman W7pF TekScopes Moderator |
|
Thanks for all those they weighed in on the TD issue. I was able to get back in to the lab and retest the TD this evening using Dennis¡¯s settings for the curve tracer. After lots of reading and thinking today I have learned a few things.
1) I was WAY overly skittish given the rarity of the part and partly for my lack of knowledge I did not want to hook the diode up to a power supply directly especially this one as replacement are rare and very expensive so if it was good I did not want to fry it. 2) My measurement setup was FAR from ideal. No there were long leads ALL over the bench. But I had ran collector -> amp in -> meter -> amp out -> to TD+ -> TD - -> emitter on the 577. The really high resistance value was done before I asked a question here on the group and was selected for my being very skittish with this TD also was a problem. 3) The message from this morning was to try and clarify my post from last night no additional testing was preformed this morning. 4) The oscillation noted from the photo this morning. I was so intent that I did not notice what was going on it was caused by the DMM switching ranges and what I am assuming shunt resisters as well 5) In NPN mode the curve tracer is NOT DC it is pulsed dc the meter would never be able to accurately measure the dc level output from the curve tracer. Trust the curve tracer to do its job/ Learnings Let the curve tracer do its job. It is a wonderful instrument that I need to trust and not try and second guess it. It will give me an accurate picture as to what is going on. There will be 2 pictures posted after this The good news is the diode seems to be in wonderful shape and I am now getting a curve that I expect out of them. Most of my errors this time were between instruments and chair. Knowing that the diode is good at this point leads me down a path that the diode is not getting enough drive so it seems I have a current issue on the 284. I will endeavor to trust my self and my instrumentation more in the future. /g/TekScopes/album?id=248216 Eric |
|
YIKES!!! I did not catch that part about the ammeter in your first email.
toggle quoted message
Show quoted text
DO NOT USE EXTERNAL AMMETERS. DO NOT CONNECT ANYTHING ELSE EXCEPT THE TD TO THE CURVE TRACER. That screws up everything you are trying to do. That explains the weird results on the picture you took with the two straight lines. I said they looked like resistors. NOW I KNOW WHY - THEY WERE RESISTORS in the ammeter. For the curve tracer to work properly it must be the only thing connected to your TD. Do not even use a voltmeter. Remove EVERYTHING except the curve tracer and the TD. #1) Only after you have done that, and you have followed the setup I gave you to the letter (with no additions or deviations), take a picture and post it. #2) Take a second picture and post it as well. This should be of the TD hooked up to the curve tracer displaying the response on the screen so we can see your entire setup including the TD, the connection to the CT, and the image on the CRT. Only then can we be sure you aren't creating your own problems. Dennis Tillman W7pF -----Original Message-----
From: [email protected] [mailto:[email protected]] On Behalf Of Eric Sent: Sunday, June 07, 2020 6:59 PM To: [email protected] Subject: Re: [TekScopes] Infamous D180 is it good or bad. Thanks for all those they weighed in on the TD issue. I was able to get back in to the lab and retest the TD this evening using Dennis¡¯s settings for the curve tracer. After lots of reading and thinking today I have learned a few things. 1) I was WAY overly skittish given the rarity of the part and partly for my lack of knowledge I did not want to hook the diode up to a power supply directly especially this one as replacement are rare and very expensive so if it was good I did not want to fry it. 2) My measurement setup was FAR from ideal. No there were long leads ALL over the bench. But I had ran collector -> amp in -> meter -> amp out -> to TD+ -> TD - -> emitter on the 577. The really high resistance value was done before I asked a question here on the group and was selected for my being very skittish with this TD also was a problem. 3) The message from this morning was to try and clarify my post from last night no additional testing was preformed this morning. 4) The oscillation noted from the photo this morning. I was so intent that I did not notice what was going on it was caused by the DMM switching ranges and what I am assuming shunt resisters as well 5) In NPN mode the curve tracer is NOT DC it is pulsed dc the meter would never be able to accurately measure the dc level output from the curve tracer. Trust the curve tracer to do its job/ Learnings Let the curve tracer do its job. It is a wonderful instrument that I need to trust and not try and second guess it. It will give me an accurate picture as to what is going on. There will be 2 pictures posted after this The good news is the diode seems to be in wonderful shape and I am now getting a curve that I expect out of them. Most of my errors this time were between instruments and chair. Knowing that the diode is good at this point leads me down a path that the diode is not getting enough drive so it seems I have a current issue on the 284. I will endeavor to trust my self and my instrumentation more in the future. /g/TekScopes/album?id=248216 Eric -- Dennis Tillman W7pF TekScopes Moderator |
|
Hi Eric,
toggle quoted message
Show quoted text
These new pictures look very nice and can compete with that of Raymond! One suggestion, please add the vertical sensitivity to the text under the pictures, just for future reference. I think you followed Dennis' advice, 2 mA/div. If your 577 is still in spec then Ip is just below 20 mA. Normally the Bias range of the 284 should be sufficient then for proper operation. But there is more to consider besides Ip drift. Bias current is supplied via R180, inside the tube. That's a cc resistor, prone to drift up in resistance when aging. That reduces the bias current. If you disconnect P180 then you can measure R180 between J180 and Pulse Out. (For highest protection of the TD you might first connect center Pulse Out to GND.) Note that there have been diagram changes in the bias circuits. Albert On Mon, Jun 8, 2020 at 03:58 AM, Eric wrote:
|
(continued)(once more continued, sorry)Bias current is supplied via R180... and from the other side via R184B+C which can be measured in a similar R184B+C is for symmetry in bias, to prevent a pre-pulse voltage ramp across R184A. But a drift in R184B+C or A has far less effect on the TD bias current than a same percentage drift in R180. Albert. |
|
Hi Eric,
toggle quoted message
Show quoted text
Excellent results. Perfect TD curves. Nice Photos. You have a working TD. Dennis Tillman W7pF -----Original Message-----
From: [email protected] [mailto:[email protected]] On Behalf Of Albert Otten Sent: Monday, June 08, 2020 2:02 AM To: [email protected] Subject: Re: [TekScopes] Infamous D180 is it good or bad. Hi Eric, These new pictures look very nice and can compete with that of Raymond! One suggestion, please add the vertical sensitivity to the text under the pictures, just for future reference. I think you followed Dennis' advice, 2 mA/div. If your 577 is still in spec then Ip is just below 20 mA. Normally the Bias range of the 284 should be sufficient then for proper operation. But there is more to consider besides Ip drift. Bias current is supplied via R180, inside the tube. That's a cc resistor, prone to drift up in resistance when aging. That reduces the bias current. If you disconnect P180 then you can measure R180 between J180 and Pulse Out. (For highest protection of the TD you might first connect center Pulse Out to GND.) Note that there have been diagram changes in the bias circuits. Albert On Mon, Jun 8, 2020 at 03:58 AM, Eric wrote:
-- Dennis Tillman W7pF TekScopes Moderator |
|
Today, I put two 10 mA TD's on my 576 CT. I set it at 15 V max, 650 Ohm series resistor (lowest setting of 576).
Nothing special, NPN polarity, internal Step Generator off. TD between "C" and "E". Nice curves (see two "TD 10 mA" images), see two added images. I don't understand why you (Albert and Dennis), are surprised by the clarity of the curves (Eric's and mine). Maybe you used a 577 and the 576 is an exception. TD 10 mA - 2 fell ill while on the CT and hasn't recovered. See text accompanying image in album. Apologies for the lesser quality photographs. There was a lot of light behind me that I was trying to screen with my body. Raymond |
I don't understand why you (Albert and Dennis), are surprised by the clarityHi Raymond, It is what I thought to remember. Perhaps that was more related to displaying the negative resistance part of the IV-curve. Today I also did some 576 measurements and these gave nice results like in your pictures. One measurement was with the 284, P181 and P184 disconnected. Pulse Out fed from the 576, range 15V / 650R. This way of course the resistance of R184A (54R nominally) largely determined the slope of the ascending and descending parts, but reading Ip was no problem (also somewhat below 20 mA.) The other measurement was with a TD716, 4.7 mA type. Mounted directly in a 576 diode adapter. Again a perfect loop and a very clean trace. Albert |
Raymond Domp Frank
to navigate to use esc to dismiss