I've just started using an IT-18 to test high beta small signal transistors and have noticed a few limitations.
It tests at 1.5V Vce, but it is interesting that the cal and measurements are current based so after a quick look at the schematics it seems that I could just add a 9V battery connector to get close to the usual 10V spec. It would no longer work in circuit so a push button for the higher voltage might make sense.
2. High, hfe even on the X10 scale ends up almost full scale where 500 to 1000 is
just about .25" on the scale - hard to even read to the 100s correctly. It would be
nice to have a X100 to bring it back to mid scale.
3. It tests at 1mA, would be nice to have 10 and 100 mA for larger devices.
4. There is very little current limiting on Ib seems there should be a larger fixed
resistor in the path.
Anyone done such mods?
I should just buy something that does these tests - suggestions?
It tests at 1.5V Vce, but it is interesting that the cal and measurements are current based so after a quick look at the schematics it seems that I could just add a 9V battery connector to get close to the usual 10V spec. It would no longer work in circuit so a push button for the higher voltage might make sense.
2. High, hfe even on the X10 scale ends up almost full scale where 500 to 1000 is
just about .25" on the scale - hard to even read to the 100s correctly. It would be
nice to have a X100 to bring it back to mid scale.
3. It tests at 1mA, would be nice to have 10 and 100 mA for larger devices.
4. There is very little current limiting on Ib seems there should be a larger fixed
resistor in the path.
Anyone done such mods?
I should just buy something that does these tests - suggestions?
I usually build dedicated fixtures for transistor sorting. Testers like the Heathkit are great for finding bad transistors but not real good for sorting or quantifying transistors. A curve tracer is the best available tool (source-measure units are too expensive). A fixture need not be complex, a transistor socket and a few R's and C's + a power supply will usually do the job. Set it up for your target operating point.
For less than you would spend on a socket there are a jillion really good transistor testers that can do a basic job of sorting parts on eBay. Search for Transistor tester transistor tester | eBay More here http://www.eevblog.com/forum/testgear/$20-lcr-esr-transistor-checker-project/ go to the end.
For less than you would spend on a socket there are a jillion really good transistor testers that can do a basic job of sorting parts on eBay. Search for Transistor tester transistor tester | eBay More here http://www.eevblog.com/forum/testgear/$20-lcr-esr-transistor-checker-project/ go to the end.
Look here at post 323:
http://www.diyaudio.com/forums/solid-state/286742-yet-another-adcom-gfa-565-thread-33.html
http://www.diyaudio.com/forums/solid-state/286742-yet-another-adcom-gfa-565-thread-33.html
I've built jigs before, I just want to be able to get an idea of beta per data sheet spec
to check when I buy new parts.
I tried the higher supply in place of the battery and there is just the simple problem of the 500K Ib
cal pot not having enough range. Going roughly 10X on voltage, obviously would require a 5M pot,
I could add a 6 pos coarse switch with 1M per step.
For now I pulled the collector lead to the socket and brought out a twisted pair that is
hooked up to my floating bench supply, add 8.5 V to get 10 Vce - this works. You have to
remember to reverse the leads for PNP/NPN, would be nice to have one more wafer on
the switch to reverse automatically. A 9V battery provides 10.5 which is also a simple
option.
Next I wanted to up the Ic and decided first to put in a transistor with a 10R resistor in
the collector lead, what do you know 3.87 mA stock when the manual says 1mA. Checked
it again with DVM on mA and the same thing.
Perhaps the 1mA was a for example?
Perhaps the meter changed?
The 11.8 ohm precision resistor measures fine.
Measured beta is in the ballpark.
It seems to work fine, but I'll check the wiring.
The meter is 200uA full scale and 225 R so:
Vfull scale = 200uA * 225 = .0045V
Collector current shunt is 11.8 R so:
Ic = .0045/11.8 = 3.8 mA plus .2 for the meter = 4.0 mA nice round number.
Correct R for 1mA:
R = .0045/.0008 = 5.63 R
For 10 mA:
R = .0045/.0098 = .46R
Looks like a design error.
to check when I buy new parts.
I tried the higher supply in place of the battery and there is just the simple problem of the 500K Ib
cal pot not having enough range. Going roughly 10X on voltage, obviously would require a 5M pot,
I could add a 6 pos coarse switch with 1M per step.
For now I pulled the collector lead to the socket and brought out a twisted pair that is
hooked up to my floating bench supply, add 8.5 V to get 10 Vce - this works. You have to
remember to reverse the leads for PNP/NPN, would be nice to have one more wafer on
the switch to reverse automatically. A 9V battery provides 10.5 which is also a simple
option.
Next I wanted to up the Ic and decided first to put in a transistor with a 10R resistor in
the collector lead, what do you know 3.87 mA stock when the manual says 1mA. Checked
it again with DVM on mA and the same thing.
Perhaps the 1mA was a for example?
Perhaps the meter changed?
The 11.8 ohm precision resistor measures fine.
Measured beta is in the ballpark.
It seems to work fine, but I'll check the wiring.
The meter is 200uA full scale and 225 R so:
Vfull scale = 200uA * 225 = .0045V
Collector current shunt is 11.8 R so:
Ic = .0045/11.8 = 3.8 mA plus .2 for the meter = 4.0 mA nice round number.
Correct R for 1mA:
R = .0045/.0008 = 5.63 R
For 10 mA:
R = .0045/.0098 = .46R
Looks like a design error.
Last edited:
Add two 9V batteries in series with the 1.5V batteries with a double pole switch to take out the 9V batteries most of the time. We did this with my unit 47 years ago. It works! I still use it today.
Very neat to hear that you did a similar mod.
My plan was for one, or two and to use a push button so that I would not forget and
leave it in boost high voltage. The only issue is that Ib is too large and I think a
coarse cal pot would fix that.
Right so if the boost goes in series with the existing battery, then the existing PNP/NPN
switch handles the reversal.
The nice thing about the external supply is that it has a current limiter, and I could go
up to 50V for this particular variable supply.
So I made a few mistakes here, should be:
The meter is 200uA full scale and 225 R so:
Vfull scale = 200uA * 225 = .045V
Collector current shunt is 11.8 R so:
Ic = .045/11.8 = 3.8 mA plus .2 for the meter = 4.0 mA nice round number.
Correct R for 1mA:
R = .045/.0008 = 56.3 R
For 10 mA:
R = .045/.0098 = 4.6R
EDIT: JUST RE-READ THE THEORY OF OPERATION AND IT IS DESIGNED FOR 4mA, NOT
SURE WHERE I GOT THE 1mA idea.
The meter is 200uA full scale and 225 R so:
Vfull scale = 200uA * 225 = .045V
Collector current shunt is 11.8 R so:
Ic = .045/11.8 = 3.8 mA plus .2 for the meter = 4.0 mA nice round number.
Correct R for 1mA:
R = .045/.0008 = 56.3 R
For 10 mA:
R = .045/.0098 = 4.6R
EDIT: JUST RE-READ THE THEORY OF OPERATION AND IT IS DESIGNED FOR 4mA, NOT
SURE WHERE I GOT THE 1mA idea.
Last edited:
I'll keep an eye out for a Heath IM-30, good that it will do power devices:
Heathkit Virtual Museum | IM-30
Heathkit Virtual Museum | IM-30
Now I remember why I hate this tester
Looking at the schematics at a glance I thought years ago, it has current limiting
in the base drive. That is not really true, the resistors are the current shunts.
With a 1.5V battery, lets call the Vbe roughly .5V then there is 1 V across the
current shunts. With 225 R in the beta X10 position and 25 R in the X1:
Beta on 10X: Ib = 1/225 = 4.4mA
Beta on 1X: Ib = 1/25 = 40 mA
I'm pretty sure that 40 mA Ib will destroy one of the modern 50 mA Ic max devices.
Pretty sure I just killed an old 2N2222, a nice NOS Motorola in a metal can. I know that
it is an NPN but I had the meter set to PNP, cranked up the cal got nothing switched
to off then back to NPN forgetting to turn down the cal. Also possible that with 40mA
Ib, Ic tried to go to 4A and that killed it.
The operator's manual has no mention of turning down the Cal before switching it on.
The meter makes quite the dive when you have it set to the wrong type, crank it
all the way up, then switch to the correct type, LOL. I'm too absent minded to
remember to reset things. Also, the meter goes full scale with .045 V even a Germanium
protection diode that they used does little to protect the meter.
After using Tektronix and HP gear you really have to hate Heathkit, sorry if you like it.
Good price, that's about it.
Looking at the schematics at a glance I thought years ago, it has current limiting
in the base drive. That is not really true, the resistors are the current shunts.
With a 1.5V battery, lets call the Vbe roughly .5V then there is 1 V across the
current shunts. With 225 R in the beta X10 position and 25 R in the X1:
Beta on 10X: Ib = 1/225 = 4.4mA
Beta on 1X: Ib = 1/25 = 40 mA
I'm pretty sure that 40 mA Ib will destroy one of the modern 50 mA Ic max devices.
Pretty sure I just killed an old 2N2222, a nice NOS Motorola in a metal can. I know that
it is an NPN but I had the meter set to PNP, cranked up the cal got nothing switched
to off then back to NPN forgetting to turn down the cal. Also possible that with 40mA
Ib, Ic tried to go to 4A and that killed it.
The operator's manual has no mention of turning down the Cal before switching it on.
The meter makes quite the dive when you have it set to the wrong type, crank it
all the way up, then switch to the correct type, LOL. I'm too absent minded to
remember to reset things. Also, the meter goes full scale with .045 V even a Germanium
protection diode that they used does little to protect the meter.
After using Tektronix and HP gear you really have to hate Heathkit, sorry if you like it.
Good price, that's about it.
Last edited:
I've completely modded this tester.
Added a 3 pos Ic switch with 1, 10, 100 mA. It switches the current shunt in the collector
lead and it also switches a current limiter resistor in the base lead, 10K, 1K and 100 ohms for:
.1 mA on 1mA Ic
1mA on 10mA Ic
10mA on 100mA Ic (should probably drop this to 5mA for a bit more protection)
The 1 mA Ic position is fairly useless after this mod, but I left it in case I ever want to mod it
down the road. The meter is not sensitive enough to make it work but jacks could be added
to use a DMM in place of the meter.
You have to add a shunt to the meter to calibrate Ib and beta to the new 10mA Ic main
scale, then another new shunt for the X1 scale.
All of the Ic shunts have to change.
I did not change the Ib shunts with scale changing, this is how it works:
Ic setting Beta settings
1mA X1 is really X.1, X10 is really X1 - Can only read up to Beta = 100 almost useless
10mA X1 and X10 read as indicated
100mA X1 is actually X10, X10 is actually X100
I tested mostly with a med power TO220 transistor but it had a Beta of 240.
Works fine, can't read it on the 1mA scale, others are fine.
2N2222 measures fine, Beta is 140 can't read it on the 1mA scale.
Measured some TO3 power transistors from the junk box:
2N5302 10ma Beta = 40, 100mA beta = 60
2N5885 10mA Beta = 80, 100mA beta = 95
Seems to work fine, did not finish the 10V Vce mod.
Just realized that with high beta transistors, even 1 mA Ib limit is enough to destroy
a 100mA Ic rated transistor. Perhaps a 50mA fuse in series with the 100mA shunt
and a 250mA in series with the 100 mA shunt, trying to keep it simple considering it is
dual polarity for NPN/PNP. An electronic current limit could go on the battery side
before the reversal switch so as to not need to change with polarity.
I should buy a curve tracer.
Added a 3 pos Ic switch with 1, 10, 100 mA. It switches the current shunt in the collector
lead and it also switches a current limiter resistor in the base lead, 10K, 1K and 100 ohms for:
.1 mA on 1mA Ic
1mA on 10mA Ic
10mA on 100mA Ic (should probably drop this to 5mA for a bit more protection)
The 1 mA Ic position is fairly useless after this mod, but I left it in case I ever want to mod it
down the road. The meter is not sensitive enough to make it work but jacks could be added
to use a DMM in place of the meter.
You have to add a shunt to the meter to calibrate Ib and beta to the new 10mA Ic main
scale, then another new shunt for the X1 scale.
All of the Ic shunts have to change.
I did not change the Ib shunts with scale changing, this is how it works:
Ic setting Beta settings
1mA X1 is really X.1, X10 is really X1 - Can only read up to Beta = 100 almost useless
10mA X1 and X10 read as indicated
100mA X1 is actually X10, X10 is actually X100
I tested mostly with a med power TO220 transistor but it had a Beta of 240.
Works fine, can't read it on the 1mA scale, others are fine.
2N2222 measures fine, Beta is 140 can't read it on the 1mA scale.
Measured some TO3 power transistors from the junk box:
2N5302 10ma Beta = 40, 100mA beta = 60
2N5885 10mA Beta = 80, 100mA beta = 95
Seems to work fine, did not finish the 10V Vce mod.
Just realized that with high beta transistors, even 1 mA Ib limit is enough to destroy
a 100mA Ic rated transistor. Perhaps a 50mA fuse in series with the 100mA shunt
and a 250mA in series with the 100 mA shunt, trying to keep it simple considering it is
dual polarity for NPN/PNP. An electronic current limit could go on the battery side
before the reversal switch so as to not need to change with polarity.
I should buy a curve tracer.
Last edited:
Slept on it and I have the panel jacks for meter probe tips and all that's needed is
to probe across the decade value Ib current limiting resistors. The meter is
auto ranging, accurate and it is then easy to read beta >200 without a need for
a beta switch. The 1mA and 10mA settings are then much more useful.
I have 16VDC 5A laptop power adapters, hmmm, high power LM317 regulator with
current limiting set by the Ic selector. The switch has room for 1A and 5A Ic settings.
to probe across the decade value Ib current limiting resistors. The meter is
auto ranging, accurate and it is then easy to read beta >200 without a need for
a beta switch. The 1mA and 10mA settings are then much more useful.
I have 16VDC 5A laptop power adapters, hmmm, high power LM317 regulator with
current limiting set by the Ic selector. The switch has room for 1A and 5A Ic settings.
Greetings!
I had been following the Adcom PA repair thread where this HK unit was being used and was considering purchase of one. I already have a Peak DCA55. are there any tests that the HK unit would be a better device to use? I typically use the Peak to verify lead outs (very handy on old devices), test gain and leakage. I found a HK on 'bay for a bit over $40 US, but suspect it would need some basic maintenance and perhaps some of the upgrades mentioned in this thread.
Thanks for any tips!
Best, Jim
I had been following the Adcom PA repair thread where this HK unit was being used and was considering purchase of one. I already have a Peak DCA55. are there any tests that the HK unit would be a better device to use? I typically use the Peak to verify lead outs (very handy on old devices), test gain and leakage. I found a HK on 'bay for a bit over $40 US, but suspect it would need some basic maintenance and perhaps some of the upgrades mentioned in this thread.
Thanks for any tips!
Best, Jim
Just noticed that the Heathkit IT-3120 is similar to the IT-18 but has Ic current
selection and several more features:
http://lcweb2.loc.gov/master/mbrs/r...anual Model IT-3120 FET-Transister Tester.pdf
But the IM-36 looks even better:
http://tubularelectronics.com/Heath...Manuals_IM-IN/IM-36/Heath IM-36 Manual_v6.pdf
selection and several more features:
http://lcweb2.loc.gov/master/mbrs/r...anual Model IT-3120 FET-Transister Tester.pdf
But the IM-36 looks even better:
http://tubularelectronics.com/Heath...Manuals_IM-IN/IM-36/Heath IM-36 Manual_v6.pdf
- Status
- Not open for further replies.