PopularFX
Home Help Search Login Register
Welcome,Guest. Please login or register.
2021-12-07, 21:51:05
News: Forum TIP:
The SHOUT BOX deletes messages after 3 hours. It is NOT meant to have lengthy conversations in. Use the Chat feature instead.

Pages: 1 ... 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 [44] 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 ... 103
Author Topic: Dally, Shark & Ruslan workbench  (Read 76288 times)

Group: Professor
Hero Member
*****

Posts: 2602
To accurately measure the Vbe, simply place one probe at the emitter and another probe at the base where these leads exit the device and measure the differential.  This eliminates all lead inductances.
Or ditch the long ground clip and use a short springy one instead.
https://i.stack.imgur.com/7OPgl.jpg
Dally, Shark & Ruslan workbench


However, I still think that the RF EMI from the Kacher can enter the scope through its power supply cable and falsify the measurement ...unless Itsu's scope can be made battery powered and floated.

OTOH: Since Itsu has burned out the Q6 already, perhaps the VBE does get that high and the base needs an extra protection.
Do you think a 5W Zener diode to ground would be fast enough to protect the base of Q6 ?
   
Sr. Member
****

Posts: 289
Yes, the 200mA during DC bias conditions, does not mean a 200mA amplitude of current flowing through the primary coil.

I agree with everything PW wrote above.

@PW
Since Q6 is recommended for audio amplifier applications, I wonder whether its base-emitter parasitic capacitance (Cπ) and Drain-Source capacitance (CDS) of the Q4 are large enough to limit base currents at the high frequencies arriving from the coil/antenna.

Also, do you think a 4.2V 5W Zener diode to ground would be fast enough to protect the base of Q6 ?

Verpies,

Regardless of any Q4/Q6 capacitance, if the observed base excursions are accurate and not just ground or inductance related artifacts, significant base voltages/ base currents are inferred (as well as collector currents).

Standard zener diodes are probably too slow for clamping the base.  Reverse recovery time or junction capacitance is rarely provided on zener data sheets.  There are faster clamping devices available that are "zener like", i.e., transorbs, etc.

However, because we are only concerned about AC/pulsed excursions on the base, and not DC conditions, fast, low capacitance diodes could be used, reverse biased to the desired clamp voltage, with the bias voltage well bypassed with HF/low ESR capacitors.

Regarding any protection scheme , however, it would be most interesting to know what the base and collector currents actually are during operation.

Do you recall how fast (freq response) Itsu's current probe and homemade current sensors were able to operate?

PW 
   
Group: Experimentalist
Hero Member
*****

Posts: 1380
Or ditch the long ground clip and use a short springy one instead.
https://i.stack.imgur.com/7OPgl.jpg
Dally, Shark & Ruslan workbench


However, I still think that the RF EMI from the Kacher can enter the scope through its power supply cable and falsify the measurement ...unless Itsu's scope can be made battery powered and floated.

OTOH: Since Itsu has burned out the Q6 already, perhaps the VBE does get that high and the base needs an extra protection.
Do you think a 5W Zener diode to ground would be fast enough to protect the base of Q6 ?

I agree with the RF EMI possibly entering the probes and the RF ground probe connection would be a good thing to try with a single probe.

Not sure about the zener being fast enough.  Perhaps multiple (3-4) series connected Schottky diodes.

Regards,
Pm
   
Sr. Member
****

Posts: 289
Again, the Vbe of a bipolar power transistor such as the 2SC5200 can in no way reach 5v or anything close.  The actual max Vbe encountered should be ~1.5v at the maximum collector current.  In Itsu's scope shots of the base voltage is not an accurate measurement of the actual base voltage IMO.

You are very likely correct regarding your above comments, and what we are seeing "may" not be accurate.  However, some voltage or current spec was apparently exceeded when Itsu attempted 100 volt operation.  As well, there are base voltage excursions indicated that are negative with respect to the emitter , i.e., reverse biased, that might easily exceed 5 volts at higher collector voltages.

Quote
To accurately measure the Vbe, simply place one probe at the emitter and another probe at the base where these leads exit the device and measure the differential.  This eliminates all lead inductances.

Regards,
Pm

A differential measurement, as you suggest, is an excellent idea and may provide more accurate data.

PW
   
Group: Experimentalist
Hero Member
*****

Posts: 1380
You are very likely correct regarding your above comments, and what we are seeing "may" not be accurate.  However, some voltage or current spec was apparently exceeded when Itsu attempted 100 volt operation.  As well, there are base voltage excursions indicated that are negative with respect to the emitter , i.e., reverse biased, that might easily exceed 5 volts at higher collector voltages.

A differential measurement, as you suggest, is an excellent idea and may provide more accurate data.

PW

Several years ago I spoke to a Tek engineer about the MDO3000 series scope and he said that two pair of the four probes shared a common ground on the mother board ie, CH1 and CH3/CH2 and CH4 but it could have been CH1 and CH2/CH3 and CH4 as I don't remember for sure.  In any case, each pair of probes will share a common ground and I think that should result in a common mode rejection of the ground lead induced EMI.

Regards,
Pm
   
Sr. Member
****

Posts: 289
Or ditch the long ground clip and use a short springy one instead.
https://i.stack.imgur.com/7OPgl.jpg
Dally, Shark & Ruslan workbench


However, I still think that the RF EMI from the Kacher can enter the scope through its power supply cable and falsify the measurement ...unless Itsu's scope can be made battery powered and floated.

OTOH: Since Itsu has burned out the Q6 already, perhaps the VBE does get that high and the base needs an extra protection.
Do you think a 5W Zener diode to ground would be fast enough to protect the base of Q6 ?

Verpies,

I have a real love/hate relationship with those probe tips.  They are a pain to use handheld in tight confines, particularly if using more than one probe set-up like that.  But perhaps a Vbe measurement made in such a manner would provide a useful comparison.  Equipment (scope, supplies) related ground loops/currents are still a possible source of error.  I always short the two tips to a trace/groundplane at or near my measurement point to see what my "zero voltage" condition actually looks like.

I have on several occasions used the machined contacts removed from a DIP socket as a connector to to actually plug the probe tip and ground lead into.  They are quite small and can be soldered to the desired testpoints.  I usually tack them where I want to measure in an almost horizontal orientation so that it doesn't take much to support the probe handle with "something" for hands free operation.

PW   
   

Group: Professor
Hero Member
*****

Posts: 2602
In any case, each pair of probes will share a common ground and I think that should result in a common mode rejection of the ground lead induced EMI.
And is that "ground lead induced EMI" smaller or larger when the ground leads are left floating at the probe ends.
Do you recommend connecting them there together and/or to the emitter?
   

Group: Professor
Hero Member
*****

Posts: 2602
I have a real love/hate relationship with those probe tips.  They are a pain to use handheld in tight confines, particularly if using more than one probe set-up like that. 
I feel about them the same way ...so I just solder the springy ground tips to my test ground points (only the tip - not the spiral).  I use disposable plain copper wire for them so they solder well but are not really springy.

If I can't solder them (due to small SMD parts or pads), then I bend them, manipulate them in place, get intermittent contact and get annoyed.
   

Group: Elite Experimentalist
Hero Member
*****

Posts: 3270
Itsu,

Be aware that if the "200ma through the transistor" you mention is your collector current as measured during DC bias conditions, that 200ma will have little meaning under dynamic conditions.

While operating at 45 volts, the base is seeing voltages (spikes) that approach 3.5v.  Whether that base voltage is "real" or an artifact moreso related to ground or lead inductance needs to be determined.  However, if the base is indeed being driven to ca 3.5v, that infers very high base peak currents.  As well, the absolute maximum for Vbe is 5v.  It is likely that adjusting the 45v collector supply much higher will start putting those base excursions very close to that limit.

Q6 is saturating so depending on the load impedance, Ic could be approaching a very significant portion of the maximum allowable collector current.

When probing Vbe and Vce, attach your scope probe ground lead directly on the emitter or as close to it as possible. 

Perhaps discussions should shift toward possible ways to measure Ib and Ic under dynamic conditions with minimal effect on circuit operation.

How fast are your current probes?


PW



   


The 200mA i mentioned was a reference to myself when adjusting the collector voltage.
I used a 0 to 300V or so PS to slowly increase the voltage making sure the current delivered by this PS stayed around 200mA while adjusting R17.
Its an analoge 1A current meter.

I will look into the base signal spikes and try to use the mentioned differential probing methode and/or the HF probe tips to see if they are real or artifacts

My current probe is the A6302 which is a DC to 50 MHz current probe, but its defective at the moment and a just received replacement part seems defective also.

I currently only have a P6021 AC current probe which is specified for 60MHz.

But doing current measurements using such a probe would mean to extend the base.


Itsu
   
Sr. Member
****

Posts: 289



I currently only have a P6021 AC current probe which is specified for 60MHz.

But doing current measurements using such a probe would mean to extend the base.


Itsu

But you could use that probe to look at collector current and the current at the antenna input (J2).

PW
   

Group: Elite Experimentalist
Hero Member
*****

Posts: 3270

Yes, i probably can (using my old OWON scope which has limited features), let me try.



Below screenshot is the base signal (12V on collector, oscillations going) with both a HF probe tip (red) and the normal probe tip / ground lead) yellow.

Seems there is little difference there.

Itsu
   
Group: Experimentalist
Hero Member
*****

Posts: 1380
And is that "ground lead induced EMI" smaller or larger when the ground leads are left floating at the probe ends.
Do you recommend connecting them there together and/or to the emitter?

I would recommend floating the ground leads of the probe pair as the best option.  I realized with my own frustration measuring low to medium frequencies I mean, what good does it do to have a scope with a 500Mhz bandwidth when the typical probe won't accurately measure 10's of Mhz?

For example, I made my own low impedance probe for RF CSR measurements realizing that the internal hi impedance resistors added to the signal error.  I do not have an active probe but logically they should do better in these conditions but not sure!

Regards,
Pm
   

Group: Professor
Hero Member
*****

Posts: 2602
I would recommend floating the ground leads of the probe pair as the best option. 
Floating and connected together or floating separately ?
   

Group: Elite Experimentalist
Hero Member
*****

Posts: 3270
But you could use that probe to look at collector current and the current at the antenna input (J2).

PW

1st screenshot is red collector voltage, yellow collector current (current probe set at 10mA/1mV).

2nd screenshot is red collector voltage, yellow antenna input (J2) current 8same setting as above).

Itsu
   

Group: Professor
Hero Member
*****

Posts: 2602
1st screenshot is red collector voltage, yellow collector current (current probe set at 10mA/1mV).
Looks like collector current stops oscillating but collector voltage continues to.
The peak collector current does not exceed 10% of the BJT's rating.

2nd screenshot is red collector voltage, yellow antenna input (J2) current 8same setting as above).
Looks like the antenna current oscillates 180º at 1.125MHz and 360mAP-P at the maximum, but the collector voltage contains higher frequencies - especially at higher amplitudes.
   

Group: Moderator
Full Member
*****

Posts: 235
Enjoy your trek through life but leave no tracks
Again, the Vbe of a bipolar power transistor such as the 2SC5200 can in no way reach 5v or anything close.  The actual max Vbe encountered should be ~1.5v at the maximum collector current.  In Itsu's scope shots of the base voltage is not an accurate measurement of the actual base voltage IMO.

To accurately measure the Vbe, simply place one probe at the emitter and another probe at the base where these leads exit the device and measure the differential.  This eliminates all lead inductances.

Regards,
Pm
You don't really get it do you ?   it simply means if you exceed the maximum BE voltage you destroy the transistor even for a micro second.
Try having no protection on EB like using a Tesla HV winding connected to the Base as feed back oscillator then try drawing arks off it, and see how long
Your BJT lasts, still don't forget there is also reverse BE voltage to worry about too.

Sil
   
Group: Experimentalist
Hero Member
*****

Posts: 1380
Floating and connected together or floating separately ?

Floating separately but with another ground from one of the other probes supplying the circuit ground connection.  Somewhere I have a video that I used to prove to Tektronix that they had a problem with the TPP0500B probes that were supplied with my MDO3034 in that the ground jacket induced voltage into the inner lead when measuring frequencies in the 2-10Mhz range.  That's when they told me about the common grounds for the probe pairs but offered no other solution!

Regards,
Pm
   
Group: Experimentalist
Hero Member
*****

Posts: 1380
You don't really get it do you ?   it simply means if you exceed the maximum BE voltage you destroy the transistor even for a micro second.
Try having no protection on EB like using a Tesla HV winding connected to the Base as feed back oscillator then try drawing arks off it, and see how long
Your BJT lasts, still don't forget there is also reverse BE voltage to worry about too.

Sil

Uuh, I guess I don't! :-[  Sorry!

Regards,
Pm
   
Sr. Member
****

Posts: 289
1st screenshot is red collector voltage, yellow collector current (current probe set at 10mA/1mV).

2nd screenshot is red collector voltage, yellow antenna input (J2) current 8same setting as above).

Itsu

Itsu,

What collector coil voltage were you applying in these captures (ie, 12volts)?

Regarding the traces showing collector voltage, where are you connecting your "collector voltage" probe and ground lead in these captures?

PW
   

Group: Professor
Hero Member
*****

Posts: 2602
Floating separately but with another ground from one of the other probes supplying the circuit ground connection.
i think the Owon scope has only 2 channels.

Somewhere I have a video that I used to prove to Tektronix that they had a problem with the TPP0500B probes that were supplied with my MDO3034 in that the ground jacket induced voltage into the inner lead when measuring frequencies in the 2-10Mhz range. 
That is one video I'd like to see!
   

Group: Elite Experimentalist
Hero Member
*****

Posts: 3270
Looks like collector current stops oscillating but collector voltage continues to.
The peak collector current does not exceed 10% of the BJT's rating.
Looks like the antenna current oscillates 180º at 1.125MHz and 360mAP-P at the maximum, but the collector voltage contains higher frequencies - especially at higher amplitudes.

Yes, with the current probe at 10mA/1mV, the the peak collector current is around 1.4A and the antenna current around 360mA.

The kacher/antenna combo resonates around the 1.125Mhz, the kacher secondary alone in the 2Mhz range and perhaps the primary coil has its own (higher) resonance influence.

Itsu
   

Group: Elite Experimentalist
Hero Member
*****

Posts: 3270
Itsu,

What collector coil voltage were you applying in these captures (ie, 12volts)?

Regarding the traces showing collector voltage, where are you connecting your "collector voltage" probe and ground lead in these captures?

PW

PW,

I used 12V on the collector to stay within safe limits.

I had the current probe at J2-2 for the collector current and i had the voltage probe directly across the transistor collector - emitter legs.
Same for the antenna input current.

Guess with these low voltages i can use the TDS3054B scope which has better features (differential probing etc.).


Itsu
   
Sr. Member
****

Posts: 289
Doing some experiments using a Gas Discharge Tube (GDT) of 350V at the end of the antenna to ground to see if that causes the typical kacher signal as shown in Vasik his post #1055 above.


Seems this GDT is to slow as it looks like that only "every other" gated kacher oscillation cycle is cut off.

Red is base, yellow probe near by antenna.

Itsu

Itsu,

I am not sure I understand the purpose of this test, but it does not look like the GDT is "too slow".

Note in the waveforms that  after the GDT fires, all oscillations are quenched.  On the next TC pulse, oscillations begin to build again, but do not reach sufficient voltage to fire the GDT.  At the second TC pulse period, oscillations are still decaying from the prior TC pulse providing more signal to the Q6 base and driving the oscillations to a higher voltage sufficient to fire the GDT.

What is of interest is the large spikes on the base waveform as the GDT fires.  Perhaps these spikes are moreso test lead induced or ground related artifacts than representative of the real base voltage.  However, if these spikes are indeed occurring to a large degree on the base of Q6, they are putting Q6 at risk of damage.

Additionally, there are lesser spikes seen at the end of the TC pulses where the GDT is not firing.  I had believed these spikes to be only related to Q6 turn off.  However, these spikes do appear to be aligned with the peak in voltage of the oscillations (which may also be coincidental with Q6 turn-off).

Is it possible you have some leakage occurring (capacitor breakdown, brush/corona discharge, etc) somewhere even at the lower voltage oscillations (which must be lower than the 350V of the GDT)?

PW
   
Group: Experimentalist
Hero Member
*****

Posts: 1380
i think the Owon scope has only 2 channels.
That is one video I'd like to see!

OK, finally found the short videos-

https://www.youtube.com/watch?v=wcVAbVtPpKw

https://www.youtube.com/watch?v=3fn90LAADWU

Regards,
Pm
   
Sr. Member
****

Posts: 289
1st screenshot is red collector voltage, yellow collector current (current probe set at 10mA/1mV).

2nd screenshot is red collector voltage, yellow antenna input (J2) current 8same setting as above).

Itsu

I am amazed at how clean the current waveforms are, particularly when compared to the collector voltage waveforms.

The current observed on the antenna lead seems insufficient to drive the base voltage to any voltage level that would hazard Q6.  Moreover, if the noisey/spikey waveforms seen at the base of Q6 were "real", and not the result of test lead induced or ground related artifacts, it would seem reasonable to expect to see at least some semblance of similarly noisey/spikey signals in the antenna current waveform.

Also of note is that with regard to the collector voltage waveforms is that even at the lower 12 volt operation, there appears to be voltage spikes exceeding 95 volts.  Again, if these are real voltages and not induced or ground related measurement artifacts, it seems reasonable to expect that at higher voltage operation (24v-48v), these spikes might exceed Q6's Vce breakdown voltage.

If I recall, you have a capacitor off-board across the collector power supply with a ground lead of significant length connecting that capacitor to the PCB ground.  You might try adding a parallel ground wire (of significant size) between that capacitor's ground and the PCB ground, while observing the collector voltage, to see if reducing that connection's lead inductance has any affect on the observed collector voltage waveform (ie, less noise/spike amplitude).

PW
 
   
Pages: 1 ... 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 [44] 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 ... 103
« previous next »


 

Home Help Search Login Register
Theme © PopularFX | Based on PFX Ideas! | Scripts from iScript4u 2021-12-07, 21:51:05