Series diode to reduce supply voltage

[sebmadgwick]

I have a 7.4 V battery and want to power RC servos with a max of ~6 V, the peak current might be something like 1 or 2 A; continuous 0.5 A.  The simplest solution seems to be a series diode with a forward voltage of 1.4 V.  However, finding a part is not as easy as I expected.

I have been browsing diodes but datasheets seem to only list the "maximum forward voltage".  I'm interested in the max and min forward voltage.  Something like the 1N5391 seems great but I'd like to understand a bit more about what voltage drops I might see in practice.

What do you think? Thanks in advance!

[dannyf]

Datassheets typically provide a V/I curve that provides the information you want.

The minimum voltage drop over a diode is 0.

[madires]

Silicon diodes got a voltage drop of about 0.7V. For 1.4V you can put two in series and just have to select diodes matching the current. There should be a graph showing the voltage drop vs. current in the datasheet.

[Bertho]

The forward voltage of a diode is dependent on the current. Most datasheets include a graph that show forward voltage versus forward current. The forward voltage of normal silicon diodes is between 0.6V (low current) and 1.2V (high current).

If you need a stable voltage, then a diode is not an optimal solution if the current varies. However, you normally see 0.6V..0.8V drop within reasonable bandwidth of current.

You do need to have a diode that can cope with the current. It needs to have a higher max. forward current rating than that you's need. The second thing is that you need to get away with the power dissipation. A diode with 0.7V forward voltage and 2A of current dissipates 1.4W and that becomes heat.

The diode you mention, 1N5391, has a forward voltage of 0.86V..0.95V at 0.5A (just under 0.5W dissipation at those parameters). The thermal properties are Rja = 60 deg/W, so you will see about 30 degree increase above ambient temperature if you just put it in.

[sebmadgwick]

Thanks for the replies. The graph has all the info (e.g. Fig 3 here). It looks like a couple of 1N5391 in series could do the job.  They would drop:
1.2 V at 10 mA
2.0 V at 1 A
2.2 V at 3 A (although IFmax = 1.5 A)

The supply doesn't have to be stable.  In the past I have powered HS-422 servos directly from a 7.2 V which seemed to be a common arrangement, the higher voltage results in faster motion and greater torque.  I expect the newer servos I now intend to use (HS-485HB) will behave the same.  Perhaps my design could advise that one or both the diodes be instead populated with a jumps for 'optional higher servo voltage'