[time-nuts] Re: OSA-5400 power transistor

[Bob kb8tq]

Hi

With 24V in, 450 ma is 10.8W. That’s a pretty hearty number for warming up
an OCXO. That power would not be going through the 18V regulator for a couple
of reasons. 

Once warmed up, 170 ma at 24V is 4.08W. That’s more than a functioning OCXO
should pull at typical lab bench temperatures. 60 ma at 24V is 1.44W. That is a 
reasonable power for a functioning OCXO on a lab bench. 

Bob

> On May 9, 2022, at 3:32 PM, Marek Doršic <marek.dorsic at gmail.com> wrote:
> 
> Hi Paul,
> 
>    thanks for your insight. Sadly there is no oscillations on the 18V. 
> The problem must be somewhere else. What I just do not understand is, why it starts working for couple of hours when I changed the Q4 PNP transistor. And now all voltages seems to be fine and it did not work. I also today noticed interestingly that touching the Q1 transistor metal case with a voltage probe leads to huge input current changes. While heating-up, when I touch it the current increases from 440mA to 460mA. When the unit was already heated, by touching the transistor case the input current drops from 170mA to only 60mA. I am not capable to understand the circuit right now, but for me this might be the next suspicious part.
> 
> The oven heating seems to be OK. On the PCB are pins labeled with C, B, E ( <https://www.dropbox.com/s/i6nq0fbi7p4lrin/IMG_7361.jpeg?dl=0>photo <https://www.dropbox.com/s/i6nq0fbi7p4lrin/IMG_7361.jpeg?dl=0> here) with wires soldered leading directly to the heated core. I measured voltages on these pins together with the input current. See attached graph <https://www.dropbox.com/s/37ck2mlyqgcwd46/Screenshot%202022-05-09%20at%2022.53.45.png?dl=0> below of first 40 minutes while the unit was heating up. I think this pins are the terminals of the heating transistor inside the oven and everything looks good there.  0.44A at 24V = 10.5W and then the current settles down at about 0.17A (4W). This is quite in within spec (should be 3.5W after 2.5 hours warm-up).
> 
> 
> 
>   .marek
> 
>> On 9 May 2022, at 15:54, paul swed <paulswedb at gmail.com> wrote:
>> 
>> Marek
>> Thanks. I have the schematic and can now see that its a 18V regulator. So
>> thats only 3 watts. Its a classic differential regulator so it can accept a
>> wide range of transistors because the circuit has quite a bit of gain. If
>> your transistor is being destroyed then potentially there is an oscillation
>> in the circuit.
>> A scope on the +18 should tell you.
>> Other then that the current should start high at .46 amps as you mention in
>> by 20 minutes should drop down to 46 ma as a guess. If it stays high the
>> ovens overheating and as you are concerned perhaps a bad themistor.
>> Let us know how you are doing.
>> Regards
>> Paul
>> WB8TSL
>> 
>> On Mon, May 9, 2022 at 9:11 AM Marek Doršic <marek.dorsic at gmail.com <mailto:marek.dorsic at gmail.com>> wrote:
>> 
>>> Yes, it is a power transistor with heatsing.
>>> Please find attachned the attachments via dropbox
>>> 
>>> 
>>> https://www.dropbox.com/s/efzgvs2rh8c76in/Screenshot%202022-05-08%20at%2018.58.10.png?dl=0 <https://www.dropbox.com/s/efzgvs2rh8c76in/Screenshot%202022-05-08%20at%2018.58.10.png?dl=0>
>>> <
>>> https://www.dropbox.com/s/efzgvs2rh8c76in/Screenshot%202022-05-08%20at%2018.58.10.png?dl=0 <https://www.dropbox.com/s/efzgvs2rh8c76in/Screenshot%202022-05-08%20at%2018.58.10.png?dl=0>
>>>> 
>>> 
>>> https://www.dropbox.com/s/wd4yrndn4scfzov/Screenshot%202022-05-08%20at%2019.01.46.png?dl=0 <https://www.dropbox.com/s/wd4yrndn4scfzov/Screenshot%202022-05-08%20at%2019.01.46.png?dl=0>
>>> <
>>> https://www.dropbox.com/s/wd4yrndn4scfzov/Screenshot%202022-05-08%20at%2019.01.46.png?dl=0 <https://www.dropbox.com/s/wd4yrndn4scfzov/Screenshot%202022-05-08%20at%2019.01.46.png?dl=0>
>>>> 
>>> 
>>> .marek
>>> 
>>>> On 8 May 2022, at 21:05, paul swed <paulswedb at gmail.com> wrote:
>>>> 
>>>> Marek
>>>> No diagram included that I can see.
>>>> The next comment may be totally wrong since I have nothing to go on.
>>>> If the input voltage is 24 V and the supply is 10 V reg at .48A, then
>>>> during the initial warm up the transistor easily dissipates 6 watts. That
>>>> would be a power transistor and some form of heat sink to keep the
>>> junction
>>>> temperature reasonable.
>>>> Regards
>>>> Paul
>>>> WB8TSL
>>>> 
>>>> On Sun, May 8, 2022 at 2:14 PM Marek Doršic <marek.dorsic at gmail.com
>>> <mailto:marek.dorsic at gmail.com <mailto:marek.dorsic at gmail.com>>> wrote:
>>>> 
>>>>> I would like to get your thoughts on my problem with OSA-5400
>>> oscillator.
>>>>> 
>>>>> I have on old unit, which is somehow broken. I was told it was
>>> overpowered
>>>>> with voltages up to 32V (standard supply voltage is 24V) and even
>>> sourced
>>>>> with reverse polarity supply power.
>>>>> 
>>>>> When I first powered it up, it draws only 2mA. I replaced what I
>>> supposed
>>>>> was a broken 10V voltage reference (how wrong I was), with a 10V zener
>>>>> diode and voilà , I had a nice steady 5MHz, 14dB signal. But only for
>>> couple
>>>>> of hours and then it died again. So I reverse engineered the schematics
>>>>> below and the part in question (Q4) is what I suppose a PNP power
>>>>> transistor. A bought a bunch of different types available. Solder in an
>>>>> 2N2905A and powered the unit. The heater went on, the unit drew 480 mA
>>>>> after power up but the output signal was still only some noisy 2mVp-p.
>>>>> After a few minutes the transistor went broken and the heater and
>>>>> everything went off.
>>>>> Then I put there a BC160-10. This seemed to be good choice. The unit
>>>>> worked again normally, with nice output signal, but again only couple of
>>>>> hours and then the signal was lost.
>>>>> But all the voltages at test pads remain as labeled on the PCB. The
>>>>> thermistor output pins on front panel are always 2 Ohms. This part in
>>>>> heated core of the unit is probably already broken.
>>>>> 
>>>>> Do you have please any thoughts, what can be wrong with the unit or what
>>>>> kind of power transistor should be used (Q4). The original part has gold
>>>>> plated leads and TO-39 package.
>>>>> The resistor values are only indicative measured with multimeter while
>>>>> soldered in.
>>>>> 
>>>>> .md
>>>>> 
>>>>> 
>>>>> 
>>>>> 
>>>>> 
>>>>> 
>>>>> _______________________________________________
>>>>> time-nuts mailing list -- time-nuts at lists.febo.com <mailto:time-nuts at lists.febo.com> <mailto:
>>> time-nuts at lists.febo.com <mailto:time-nuts at lists.febo.com>> -- To unsubscribe send
>>>>> an email to time-nuts-leave at lists.febo.com <mailto:time-nuts-leave at lists.febo.com> <mailto:
>>> time-nuts-leave at lists.febo.com <mailto:time-nuts-leave at lists.febo.com>>
>>>>> To unsubscribe, go to and follow the instructions there.
>>>> _______________________________________________
>>>> time-nuts mailing list -- time-nuts at lists.febo.com <mailto:time-nuts at lists.febo.com> <mailto:
>>> time-nuts at lists.febo.com <mailto:time-nuts at lists.febo.com>> -- To unsubscribe send an email to
>>> time-nuts-leave at lists.febo.com <mailto:time-nuts-leave at lists.febo.com> <mailto:time-nuts-leave at lists.febo.com <mailto:time-nuts-leave at lists.febo.com>>
>>>> To unsubscribe, go to and follow the instructions there.
>>> 
>>> _______________________________________________
>>> time-nuts mailing list -- time-nuts at lists.febo.com -- To unsubscribe send
>>> an email to time-nuts-leave at lists.febo.com
>>> To unsubscribe, go to and follow the instructions there.
>> _______________________________________________
>> time-nuts mailing list -- time-nuts at lists.febo.com -- To unsubscribe send an email to time-nuts-leave at lists.febo.com
>> To unsubscribe, go to and follow the instructions there.
> 
> _______________________________________________
> time-nuts mailing list -- time-nuts at lists.febo.com -- To unsubscribe send an email to time-nuts-leave at lists.febo.com
> To unsubscribe, go to and follow the instructions there.