Hi Bill,
Glad to be able to help. Yes a single 1N4007 is only rated at 1KV, but you can put a pair in series for a piv of 2KV, note that each one should be paralleled with a resistor of 1M/1W and ceramic cap of 0.01uF/1KV to assure that they share the voltage equally in the reversed biased (non conducting) condition.
I prefer HF4007 which are a faster and still cheap variant on the 1N4007. Honestly I have used hexfreds in several designs and could not hear one iota of difference between them and the HF4007 implemented as above..
You might get satisfactory performance with choke input, the 5AR4, and center tapped operation. Use the single remaining 6.3V winding with LT1085 and schottky rectifiers.
I think despite the limited margin on the high voltage secondaries a conventional center tapped full wave rectifier is likely to produce better load sharing between windings than a pair of bridges where the higher of the two (or the less lossy rectifier of the two) would hogg most of the load current. There of course is no issue with unwanted circulating currents as the windings are not connected directly to each other. Stay with the choke input as this considerably reduces the peak rectifier current and will reduce heating in the secondary.
On the filament regulator front I was referring to the Linear Tech LT1085 which has much lower drop out than the 317 and would allow you to eliminate the 5V winding in your filament regulator scheme as noted above and reduce dissipation in the power transformer by as much as 10W.
A 317 would work fine on the 15V winding..
Regards, Kevin
Glad to be able to help. Yes a single 1N4007 is only rated at 1KV, but you can put a pair in series for a piv of 2KV, note that each one should be paralleled with a resistor of 1M/1W and ceramic cap of 0.01uF/1KV to assure that they share the voltage equally in the reversed biased (non conducting) condition.
I prefer HF4007 which are a faster and still cheap variant on the 1N4007. Honestly I have used hexfreds in several designs and could not hear one iota of difference between them and the HF4007 implemented as above..
You might get satisfactory performance with choke input, the 5AR4, and center tapped operation. Use the single remaining 6.3V winding with LT1085 and schottky rectifiers.
I think despite the limited margin on the high voltage secondaries a conventional center tapped full wave rectifier is likely to produce better load sharing between windings than a pair of bridges where the higher of the two (or the less lossy rectifier of the two) would hogg most of the load current. There of course is no issue with unwanted circulating currents as the windings are not connected directly to each other. Stay with the choke input as this considerably reduces the peak rectifier current and will reduce heating in the secondary.
On the filament regulator front I was referring to the Linear Tech LT1085 which has much lower drop out than the 317 and would allow you to eliminate the 5V winding in your filament regulator scheme as noted above and reduce dissipation in the power transformer by as much as 10W.
A 317 would work fine on the 15V winding..
Regards, Kevin
Kevin,
Full wave rectification with a 300-0-300V@60mA will be used for the Aikido but for the 12B4A I think I will use a bridge for a single 0-380 out of 380-0-380V because it is rated at 250mA. Right or wrong? In that case 1n4007 will do the job instead of 2 connected in series. PSUII shows the peak of 1,155V.
I can't find the HF4007 in the Farnell (Australia) catelogue. I browsed through the catelogue and found BYD360 (or some similar part number that I can't remember exactly). It is a shootky diode that stands 600V with 1A current and it is rather cheap. I wonder why nobody has mentioned this one. But in both the 12B4A and the Aikido I have to connect them in series to provide enough safty margins, which means in the 12B4A I will have to use 8 of them.
I printed out the datasheets for both the LT1086 and LM1086 and I can't find difference at all! Dropout voltage, Line regulation, Load regulation, Ripple rejection, Transient response, etc, are basically identical and one does not appear to be better than the other. While price wise, the LM1086 sells for AUD$2.66 and the LT1086 sells for AUD$17.27!!! I wonder why?
Regards,
Bill
Full wave rectification with a 300-0-300V@60mA will be used for the Aikido but for the 12B4A I think I will use a bridge for a single 0-380 out of 380-0-380V because it is rated at 250mA. Right or wrong? In that case 1n4007 will do the job instead of 2 connected in series. PSUII shows the peak of 1,155V.
I can't find the HF4007 in the Farnell (Australia) catelogue. I browsed through the catelogue and found BYD360 (or some similar part number that I can't remember exactly). It is a shootky diode that stands 600V with 1A current and it is rather cheap. I wonder why nobody has mentioned this one. But in both the 12B4A and the Aikido I have to connect them in series to provide enough safty margins, which means in the 12B4A I will have to use 8 of them.
I printed out the datasheets for both the LT1086 and LM1086 and I can't find difference at all! Dropout voltage, Line regulation, Load regulation, Ripple rejection, Transient response, etc, are basically identical and one does not appear to be better than the other. While price wise, the LM1086 sells for AUD$2.66 and the LT1086 sells for AUD$17.27!!! I wonder why?
Regards,
Bill
Hi Bill,
I think the LM1086 will work just fine.. 😀
Everything else sounds pretty reasonable as well..
Kevin
I think the LM1086 will work just fine.. 😀
Everything else sounds pretty reasonable as well..
Kevin
- Status
- Not open for further replies.