I have this radio which a friend gave me to fix. When I took the covers off a little green capacitor fell out. I have no way to know where it goes as there are many intentionally empty capacitor spots on the board. There is also one tacked on the solder side. The radio had a shorted final (IRF520). I replaced both finals. The driver tested fine so I left it alone. Also replaced the 2sb827 AM regulator with a TIP36c I had on hand as it was shorted as well. Radio now transmits and receives but the TIP36c gets too hot to touch after being keyed up for 20-30 seconds. Mosfets were biased to 3.6 volts driver and 3.7 volts finals. I can't find a schematic for this board either (EPTDX2912Z). Any help would be appreciated.
-
You can now help support WorldwideDX when you shop on Amazon at no additional cost to you! Simply follow this Shop on Amazon link first and a portion of any purchase is sent to WorldwideDX to help with site costs.
Galaxy DX66V3 AM modulator transistor very hot
- Thread starter Realred
- Start date
Have you done an Idle current check?
Reason:
You replaced one part in the chain, the previous one (Final) Failed.
Playing with dice if you don't recheck the current draw on that replaced MOSFET.
It can explain "why" the heat - the current thru the part is causing this heating because of the extra draw caused by the voltage to the Finals Gate is a little too high (or Drivers Gate also - which may have put too much power into the Final - causing it to fail) - keeping it "on" for a longer duration of time (Closer to Class A than Class AB) creates a "Latching".
They may have Blocking Caps - but the current across the MOSFET Drain to Source, if the Gate voltage is too high, will act as a short across the supply - your supply to the Final and Driver - is thru that TIP36c.
A properly adjusted mA reading will also show better PEP values because of the proper cutoff region in voltage doesn't keep the device on too long or cut it off before it's done with the cycle.
IF you don't reduce the voltage - it may take out not just your new TIP36 - the Final and the Driver will go with it.
When you turn down the Gate voltage - you change the "Idle current" draw.
We have found steps to solve your issue with determining the draw by using a DVM meter installed in the power lead from the power supply and no signal (TX) into a Dummy load - just watch the DVM in mA reading - adjusting the voltage to the Gates to lowest setting your baseline - - then turn the Driver up - to get a 50mA increase - then the Final up to show 70mA max increase from that Driver level - sounds confusing but you develop a technique that works for you doing this, it's a lot easier to check than to try and pull a mirror board jumper and install clips and keep the radio stable and the clips from shorting out when they shouldn't.
Reason:
You replaced one part in the chain, the previous one (Final) Failed.
Playing with dice if you don't recheck the current draw on that replaced MOSFET.
It can explain "why" the heat - the current thru the part is causing this heating because of the extra draw caused by the voltage to the Finals Gate is a little too high (or Drivers Gate also - which may have put too much power into the Final - causing it to fail) - keeping it "on" for a longer duration of time (Closer to Class A than Class AB) creates a "Latching".
They may have Blocking Caps - but the current across the MOSFET Drain to Source, if the Gate voltage is too high, will act as a short across the supply - your supply to the Final and Driver - is thru that TIP36c.
A properly adjusted mA reading will also show better PEP values because of the proper cutoff region in voltage doesn't keep the device on too long or cut it off before it's done with the cycle.
IF you don't reduce the voltage - it may take out not just your new TIP36 - the Final and the Driver will go with it.
When you turn down the Gate voltage - you change the "Idle current" draw.
We have found steps to solve your issue with determining the draw by using a DVM meter installed in the power lead from the power supply and no signal (TX) into a Dummy load - just watch the DVM in mA reading - adjusting the voltage to the Gates to lowest setting your baseline - - then turn the Driver up - to get a 50mA increase - then the Final up to show 70mA max increase from that Driver level - sounds confusing but you develop a technique that works for you doing this, it's a lot easier to check than to try and pull a mirror board jumper and install clips and keep the radio stable and the clips from shorting out when they shouldn't.
Hoo boy - yes!
The 324mA figure - that would also need to be VERIFIED that the Final and Driver are getting voltage ALL THE TIME.
A typical Galaxy radio MAY pull 1/4 - of an amp - that's 250mA (0.25A) so you already know you're over.
The 66V is claimed to RX at 0.3A per their "owners manual" at GalaxyRadios so the 324 figure is higher - but not by much - but knowing your Board number can help sort this out.
That "Green cap" can be a disc from one of the spots they solder on the bottom side of the board - that broke loose. You also don't want the leads of that cap shorting across solder pads of places it wasn't supposed to be even close to - so we have to locate a possibly blown part due to this discovery.
It would explain it - the rise anyway - but the problem here is the Gate voltage does not need to be on in RX - only in TX - so is there a residual voltage somewhere that is turning something on?
The incredibly fast rise in current - as long as it is only in TX and cools down in RX - then the suspected area is the Final and Driver - most likely those Gates staying on too long are causing this.
So, the radio itself - to be honest, I've seen less than 180mA when just "RX" or receiving.
Hmm, something is pulling current a 2X that rate it should be.
Congrats of the fixing, but now, you have got a radio - that somehow is turned up a bit too high.
I'd start by re-adjusting - lowering - the Gate voltage of Both Driver and Final - key it up - look at the "baseline" - see if AM Power can be turned UP and still have no signal - else the RF line with have to be "broken" by unsoldering a jumper or cap to break the RF power getting sent to the Driver and Final - effectively turning off the RF side - but letting the radio still key.
You completely turn down the Gate voltage - even lift the Gate lead, if necessary, of the PART (Upstream) you're not checking to kill any chances of RF blowing up your work - an added safety margin to keep your radio from self-destruct.
Then you "turn up the AM Carrier" as in Full voltage power to the Driver and Final - The AM Power pot - so the Drain and Source potential is at its highest at no signal. As in - No Signal to the GATE - but the GATE is ready to be opened.
This helps you see the mA Draw - in TX with now AM power signal - it should be pretty low - that 0.25 Amp figure to start the TX strip (Predriver).
Breaking the connection at the Drivers Gate or just before it hits the Driver and Final - so that type of current draw it could have when RF is flowing thru it - is stopped, so you can find out how much to unlock and hinge Open, the Doors to make them swing instead of "switch".
You look at the Ammeter and with the Gate voltages as low to the Gates as possible (As in Device Turned Off) - you look at the SUM of power consumption as a value to attain in steps.
Driver - works best at 50 to 55mA "rise" - voltage to Gate just close to turning on.
Final - Single 55mA to 70mA (Max.)
Final - Dual 70mA to 100mA (even split) (Max)
The 324mA figure - that would also need to be VERIFIED that the Final and Driver are getting voltage ALL THE TIME.
A typical Galaxy radio MAY pull 1/4 - of an amp - that's 250mA (0.25A) so you already know you're over.
The 66V is claimed to RX at 0.3A per their "owners manual" at GalaxyRadios so the 324 figure is higher - but not by much - but knowing your Board number can help sort this out.
That "Green cap" can be a disc from one of the spots they solder on the bottom side of the board - that broke loose. You also don't want the leads of that cap shorting across solder pads of places it wasn't supposed to be even close to - so we have to locate a possibly blown part due to this discovery.
It would explain it - the rise anyway - but the problem here is the Gate voltage does not need to be on in RX - only in TX - so is there a residual voltage somewhere that is turning something on?
The incredibly fast rise in current - as long as it is only in TX and cools down in RX - then the suspected area is the Final and Driver - most likely those Gates staying on too long are causing this.
So, the radio itself - to be honest, I've seen less than 180mA when just "RX" or receiving.
Hmm, something is pulling current a 2X that rate it should be.
Congrats of the fixing, but now, you have got a radio - that somehow is turned up a bit too high.
I'd start by re-adjusting - lowering - the Gate voltage of Both Driver and Final - key it up - look at the "baseline" - see if AM Power can be turned UP and still have no signal - else the RF line with have to be "broken" by unsoldering a jumper or cap to break the RF power getting sent to the Driver and Final - effectively turning off the RF side - but letting the radio still key.
You completely turn down the Gate voltage - even lift the Gate lead, if necessary, of the PART (Upstream) you're not checking to kill any chances of RF blowing up your work - an added safety margin to keep your radio from self-destruct.
Then you "turn up the AM Carrier" as in Full voltage power to the Driver and Final - The AM Power pot - so the Drain and Source potential is at its highest at no signal. As in - No Signal to the GATE - but the GATE is ready to be opened.
This is like sending Full Carrier - but there is no RF to send a carrier with. This step is so the turn on threshold of the driver and Final can be properly set with the highest peak voltage power the AM Regulator can push. It can be set right. The turn on voltage for the Gate of any MOSFET needs to be kept to a minimum turn on for the reason that when the AM Regulator "swings" audio - the Gate can still apply voltage to keep the part "on" even when the power load is swinging - too high it just stays on and doesn't follow the power swing rail as it could do, when the threshold voltage is at a pinch point to keep the device running linear instead of acting as a switch and just staying "on" all the time - ignoring input.
- it's this step - that most people miss when they replace any Driver or Final MOSFET - resetting voltages to devices' "just turn on" threshold - because the Device previously in there was set by - to just be on and follow the GATE changes to Drain to Source varying voltage. The next replacement part is still needing to be properly set to work like the older one did.
- When you neglect to change or adjust the new part into the home of the older one - many just turn the voltage on to a max setting they have seen others do or never even touch it - with the thinking "Ok, it works so what?"
- Its these moments of Now something is getting Hot - is when they ask questions.
- So why? Many forget the Gate is working - but to make the device run more linearly - the Gate threshold voltage setting needs to be considered - setting it too high just tells the Part to "latch on" (You left the Front Door open) - instead of utilizing the Gates' own varying signal present (Back Door needs to be opened to send the Signal thru the Front Door to get there) - so the Front Door and Back Door of the house you're working on - need to be set up so they can swing open and closed - like the older Bipolars can do better because of their "knee" of operation
This helps you see the mA Draw - in TX with now AM power signal - it should be pretty low - that 0.25 Amp figure to start the TX strip (Predriver).
Breaking the connection at the Drivers Gate or just before it hits the Driver and Final - so that type of current draw it could have when RF is flowing thru it - is stopped, so you can find out how much to unlock and hinge Open, the Doors to make them swing instead of "switch".
You look at the Ammeter and with the Gate voltages as low to the Gates as possible (As in Device Turned Off) - you look at the SUM of power consumption as a value to attain in steps.
Driver - works best at 50 to 55mA "rise" - voltage to Gate just close to turning on.
Final - Single 55mA to 70mA (Max.)
Final - Dual 70mA to 100mA (even split) (Max)
- Again, looking for lowest Gate voltage to set mA rise to show just ready to turn on.
Wow! You've given me a whole lot to absorb here. I'm a hobbyist who works on audio stuff, amps, tape decks, turntables that sort of stuff so I have basic test equipment and some knowledge of audio circuits but this RF stuff is like witchcraft or something. The big current draw and subsequent heating of the transistor is only in TX and cools down quickly in RX. The mosfets I bought may be fakes. They tested fine in my little chinese component tester but... I have since gotten some in from Tayda who I hear is pretty good. Is it worth replacing driver and finals just to see or should I do more targeted troubleshooting as you suggested. This board has the solid wire jumpers instead of the mirror board. I can de-solder those to test but I might need some step by step help there. Just looked at bias voltages again and the finals I can adjust down to zero volts but the driver pot at it's minimum (clockwise) is still 3.54 volts.
- No one is chatting at the moment.
-
-
dxBot:63Sprint has left the room.
-
dxBot:kennyjames 0151 has left the room.
-
-