Sick SG504


Eric
 

Is there a known failure mode when the high side of an SG504 goes dead? Reference and low are working as expected high is out to lunch. I have cleaned the switch and that does not seem to help.


John Bennett
 

The high side buffer amplifier (Q70) is the most likely suspect. This is a TO128 NPN transistor, Tek part number 151-0474-01. There are several suitable replacements from the cable industry, but the cost effective BFR94A works fine. Be sure you are ordering the TO128 (studded) package, as the NXP variant is an SOT23.
Also check to see if Q70 is getting drive at its base. If not, there is a small carbon composition resistor inside the high frequency cavity oscillator that sometimes cooks itself to death. I forget it's value, but if you open up the HF cavity (the one to the rear) this is easy to check and replace (with care).
Far less likely to fail are the HF series and shunt diodes, CR80, and CR155, CR158, & CR160, respectively.
After repair, the unit will need a recal, particularly the HF harmonic suppression and leveling step (you will need a spectrum analyzer).
73,
John Bennett
AE0AM


Eric
 

Thanks John I will check this out tonight. I have a spec an in the lab but it currently tops out at 40 Mhz. So it looks like I will need to go a touch faster. Ill see what I can find have any recommendations in the Spec an department. I am very much out of my depth there but I am learning as fast as I can.

Eric

-----Original Message-----
From: TekScopes@groups.io <TekScopes@groups.io> On Behalf Of John Bennett
Sent: Monday, February 8, 2021 11:36 AM
To: TekScopes@groups.io
Subject: Re: [TekScopes] Sick SG504

The high side buffer amplifier (Q70) is the most likely suspect. This is a TO128 NPN transistor, Tek part number 151-0474-01. There are several suitable replacements from the cable industry, but the cost effective BFR94A works fine. Be sure you are ordering the TO128 (studded) package, as the NXP variant is an SOT23.
Also check to see if Q70 is getting drive at its base. If not, there is a small carbon composition resistor inside the high frequency cavity oscillator that sometimes cooks itself to death. I forget it's value, but if you open up the HF cavity (the one to the rear) this is easy to check and replace (with care).
Far less likely to fail are the HF series and shunt diodes, CR80, and CR155, CR158, & CR160, respectively.
After repair, the unit will need a recal, particularly the HF harmonic suppression and leveling step (you will need a spectrum analyzer).
73,
John Bennett
AE0AM


Dick
 

Look at the TinySA now on the market, but be wary of
the numerous clones that may or may not work.

The ones at R&L seem to be OK.

73, Dick, W1KSZ
________________________________
From: TekScopes@groups.io <TekScopes@groups.io> on behalf of Eric <ericsp@gmail.com>
Sent: Monday, February 8, 2021 10:23 AM
To: TekScopes@groups.io <TekScopes@groups.io>
Subject: Re: [TekScopes] Sick SG504

Thanks John I will check this out tonight. I have a spec an in the lab but it currently tops out at 40 Mhz. So it looks like I will need to go a touch faster. Ill see what I can find have any recommendations in the Spec an department. I am very much out of my depth there but I am learning as fast as I can.

Eric

-----Original Message-----
From: TekScopes@groups.io <TekScopes@groups.io> On Behalf Of John Bennett
Sent: Monday, February 8, 2021 11:36 AM
To: TekScopes@groups.io
Subject: Re: [TekScopes] Sick SG504

The high side buffer amplifier (Q70) is the most likely suspect. This is a TO128 NPN transistor, Tek part number 151-0474-01. There are several suitable replacements from the cable industry, but the cost effective BFR94A works fine. Be sure you are ordering the TO128 (studded) package, as the NXP variant is an SOT23.
Also check to see if Q70 is getting drive at its base. If not, there is a small carbon composition resistor inside the high frequency cavity oscillator that sometimes cooks itself to death. I forget it's value, but if you open up the HF cavity (the one to the rear) this is easy to check and replace (with care).
Far less likely to fail are the HF series and shunt diodes, CR80, and CR155, CR158, & CR160, respectively.
After repair, the unit will need a recal, particularly the HF harmonic suppression and leveling step (you will need a spectrum analyzer).
73,
John Bennett
AE0AM


John Bennett
 

Eric,
I misspoke regarding the package type for Q70 (I should have looked it up). You want the 4-tab studded package (it has two emitter tabs) referred to in various data sheets as SOT122E, SOT48, and I think TO-117 (not TO-128). In any event, the BFR94A in that form factor should work. I have also used the the Motorola MRF587 and MRF511 with some success, although there are trade offs:
Typ. fT (GHz) VCEmax (V) Max Diss (W) ICmax (ma)
BRF94A 3.5 25 3.5 300
MRF587 5.5 17 5 200
MRF511 2.1 30 5 250

I generally go with the BFR94A, which are usually easy to find on eBay.
Good luck with your project!
john


Tom Lee
 

Thank you for the very helpful information, John. One thing that's always mystified me is that the Philips/NXP BFR94A doesn't even have the same die as the "original" BFR94A in the SOT122E. The breakdwon voltages, capacitances, max IC, ft, etc. are all different. I can't remember ever encountering a similar situation where the same transistor part number is used for such different devices (I'm excluding the data-compressed SMD marking codes).

--Tom

--
Prof. Thomas H. Lee
Allen Ctr., Rm. 205
350 Jane Stanford Way
Stanford University
Stanford, CA 94305-4070
http://www-smirc.stanford.edu

On 2/8/2021 16:08, John Bennett wrote:
Eric,
I misspoke regarding the package type for Q70 (I should have looked it up). You want the 4-tab studded package (it has two emitter tabs) referred to in various data sheets as SOT122E, SOT48, and I think TO-117 (not TO-128). In any event, the BFR94A in that form factor should work. I have also used the the Motorola MRF587 and MRF511 with some success, although there are trade offs:
Typ. fT (GHz) VCEmax (V) Max Diss (W) ICmax (ma)
BRF94A 3.5 25 3.5 300
MRF587 5.5 17 5 200
MRF511 2.1 30 5 250

I generally go with the BFR94A, which are usually easy to find on eBay.
Good luck with your project!
john




Chuck Harris <cfharris@...>
 

How about 2N3055?

-Chuck Harris

Tom Lee wrote:

Thank you for the very helpful information, John. One thing that's always mystified
me is that the Philips/NXP BFR94A doesn't even have the same die as the "original"
BFR94A in the SOT122E. The breakdwon voltages, capacitances, max IC, ft, etc. are all
different. I can't remember ever encountering a similar situation where the same
transistor part number is used for such different devices (I'm excluding the
data-compressed SMD marking codes).

--Tom


Tom Lee
 

Ah, yes, great example, Chuck. Hometaxial no more. The later 3055 clones, though, at least pretended to conform to the original datasheet numbers. The BFR94A variants don't even make a half-hearted effort to do so.

-- Cheers
Tom

--
Prof. Thomas H. Lee
Allen Ctr., Rm. 205
350 Jane Stanford Way
Stanford University
Stanford, CA 94305-4070
http://www-smirc.stanford.edu

On 2/8/2021 17:16, Chuck Harris wrote:
How about 2N3055?

-Chuck Harris

Tom Lee wrote:
Thank you for the very helpful information, John. One thing that's always mystified
me is that the Philips/NXP BFR94A doesn't even have the same die as the "original"
BFR94A in the SOT122E. The breakdwon voltages, capacitances, max IC, ft, etc. are all
different. I can't remember ever encountering a similar situation where the same
transistor part number is used for such different devices (I'm excluding the
data-compressed SMD marking codes).

--Tom